Essential functions of Cpeb4 in tissue homeostasis

Carlos Maíllo Carbajo

ADVERTIMENT. La consulta d’aquesta tesi queda condicionada a l’acceptació de les següents condicions d'ús: La difusió d’aquesta tesi per mitjà del servei TDX (www.tdx.cat) i a través del Dipòsit Digital de la UB (diposit.ub.edu) ha estat autoritzada pels titulars dels drets de propietat intel·lectual únicament per a usos privats emmarcats en activitats d’investigació i docència. No s’autoritza la seva reproducció amb finalitats de lucre ni la seva difusió i posada a disposició des d’un lloc aliè al servei TDX ni al Dipòsit Digital de la UB. No s’autoritza la presentació del seu contingut en una finestra o marc aliè a TDX o al Dipòsit Digital de la UB (framing). Aquesta reserva de drets afecta tant al resum de presentació de la tesi com als seus continguts. En la utilització o cita de parts de la tesi és obligat indicar el nom de la persona autora. ADVERTENCIA. La consulta de esta tesis queda condicionada a la aceptación de las siguientes condiciones de uso: La difusión de esta tesis por medio del servicio TDR (www.tdx.cat) y a través del Repositorio Digital de la UB (diposit.ub.edu) ha sido autorizada por los titulares de los derechos de propiedad intelectual únicamente para usos privados enmarcados en actividades de investigación y docencia. No se autoriza su reproducción con finalidades de lucro ni su difusión y puesta a disposición desde un sitio ajeno al servicio TDR o al Repositorio Digital de la UB. No se autoriza la presentación de su contenido en una ventana o marco ajeno a TDR o al Repositorio Digital de la UB (framing). Esta reserva de derechos afecta tanto al resumen de presentación de la tesis como a sus contenidos. En la utilización o cita de partes de la tesis es obligado indicar el nombre de la persona autora. WARNING. On having consulted this thesis you’re accepting the following use conditions: Spreading this thesis by the TDX (www.tdx.cat) service and by the UB Digital Repository (diposit.ub.edu) has been authorized by the titular of the intellectual property rights only for private uses placed in investigation and teaching activities. Reproduction with lucrative aims is not authorized nor its spreading and availability from a site foreign to the TDX service or to the UB Digital Repository. Introducing its content in a window or frame foreign to the TDX service or to the UB Digital Repository is not authorized (framing). Those rights affect to the presentation summary of the thesis as well as to its contents. In the using or citation of parts of the thesis it’s obliged to indicate the name of the author.

Essential functions of Cpeb4 in tissue homeostasis

Carlos Maíllo Carbajo

Barcelona, 2016

Universidad de Barcelona

IRB Barcelona

Doctorado en Biomedicina

Essential functions of Cpeb4 in tissue homeostasis

Memòria presentada per Carlos Maíllo Carbajo per optar al grau de doctor per la

Universitat de Barcelona.

Doctorando: Carlos Maíllo Carbajo

Director de la tesis: Raúl Méndez de la Iglesia

Tutor de la tesis: Antonio Zorzano Olarte

Barcelona, 2016

A mis padres, mi hermana, y a Ágata.

Index

Abbreviations������������������������������������������������11Abstract �������������������������������������������������������� 14Introduction ������������������������������������������������� 16

1� Gene expression and regulation ��������������������������������������������� 17

2� Post-transcriptional gene regulation �������������������������������������� 18

2.1 Nuclear processing .........................................................................18

2.2 Cytosolic crossroad .........................................................................21

3� The cytosolic fate of mRNAs ��������������������������������������������������� 22

3.1 mRNA decay ...................................................................................22

3.2 mRNA localization ...........................................................................22

3.3 mRNA translation ............................................................................23

4� Translational control of gene expression ������������������������������� 25

4.1 Global translational control ..............................................................25

4.2 Translation at the ER: A surveillance system ..................................26

4.3 The UPR translational regulation is mediated by uORFs ................30

4.4 mRNA-specific translational control: The RBPs ..............................32

5� Translational control by cytoplasmic polyadenylation ��������� 34

5.1 The CPEB-family of proteins ...........................................................34

5.2 The activatory and repressory functions of the CPEBs ...................39

5.3 The CPEBs in somatic tissues – CPEB4 ........................................41

6� Cellular metabolism and translation ��������������������������������������� 43

6.1 Bidirectional link between the UPR and cellular metabolism ..........43

6.2 The UPR and the etiology of metabolic disease ............................44

7� The circadian clock ������������������������������������������������������������������� 47

7.1 Post-transcriptional gene regulation by the clock ............................48

7.2 Circadian regulation of mammalian metabolism .............................48

Objectives ���������������������������������������������������� 50Methods �������������������������������������������������������� 52Results���������������������������������������������������������� 65

1. Generation of Cpeb4 ubiquitous knockout mice ...............................66

2. CPEB4 prevents the development of fatty liver disease ...................67

3. CPEB4 prevents hepatic steatosis autonomously ............................72

4. CPEB4 deletion causes mitochondrial dysfunction and defective lipid

metabolism in hepatocytes ...................................................................75

5. CPEB4 regulates the expression of ER-related proteins ..................79

6. CPEB4 is located at the ER, but does not localize mRNAs ..............81

7. CPEB4 deletion sensitizes livers to dietary fat-induced ER stress ...82

8. CPEB4 depletion leads to defective adaptation to chronic ER stress

..............................................................................................................83

9. CPEB4 protein levels are upregulated by the unfolded protein response

..............................................................................................................86

10. CPEB4 is the only CPEB family member induced by the UPR .......89

11. CPE-regulated mRNAs are activated during the UPR ....................90

12. CPEB4 mRNA levels are regulated by the molecular circadian clock

..............................................................................................................93

13. CPEB1/CPEB4 double knockout mice develop hepatosteatosis ..97

Discussion ��������������������������������������������������� 991. A novel regulatory branch of the UPR is orchestrated by CPEB4 ...100

2. CPEB4 drives the activation of CPE-containing transcripts ............101

3. The strong connection between CPEB4 and the cellular secretory

system .................................................................................................102

4. The circadian clock modulates the hepatic function of CPEB4 .......103

5. The distinctive role of CPEB4 inside the CPEB-family ....................104

6. CPEB4 protects from NAFLD development ....................................105

Conclusions ����������������������������������������������� 109References ��������������������������������������������������112Appendix ���������������������������������������������������� 133

Abbreviations

12 | ABBREVIATIONS

ALT: Alanine Transaminase

CPE: Cytoplasmic Polydenylation Element

CPEB: Cytoplasmic Polyadenylation Element Binding Proteins

CPEB4KO: CPEB4 ubiquitous knockout

CPEB4LKO: CPEB4 liver-specific knockout

EE: Energy Expenditure

ER: Endoplasmic Reticulum

FFA: Free Fatty-Acid

GSEA: Gene Set Enrichment Analysis

GWAS: Genome Wide Association studies

HFD: High-Fat Diet

H&E: Hematoxylin and Eosin

KO: Knockout

LD: Lipid Droplet

mtDNA: Mitochondrial DNA

MEF: Mouse Embryonic Fibroblast

ORC: Oxygen Consumption Rate

RER: Respiratory Exchange Ration

RIP-seq: RNA Immunoprecipitation Sequencing

ABBREVIATIONS | 13

TGA: Triglyceride

TG: Thapsigargin

TM: Tunicamycin

uORF: Upstream Open Reading Frame

UPR: Unfolded Protein Response

UTR: Unstranslated Region

VLDL: Very-Low Density Lipoprotein

WT: Wild-Type

ZT: Zeitgeber Time

Abstract

ABSTRACT | 15

The Cytoplasmic Polyadenylation Element Binding (CPEB)-family of RNA-

binding proteins is composed of four paralogs in mammals, CPEB1-4. The CPEBs

control the translation of the targeted mRNAs by modulating the length of their

poly(A) tail. This regulatory mechanism activates maternally stored mRNAs in

oocytes and drives oocyte maturation and early embryogenesis.

Genome-wide studies have revealed that the CPEBs are expressed in various

adult tissues and potentially regulate up to 20% of the vertebrate transcriptome.

However, the specific roles and regulation of each CPEB are poorly understood

beyond the embryonic stage.

In this study, we aimed to define the functions and regulation of CPEB4 in adult

tissues of mammalian organisms. For this purpose, we have generated ubiquitous and

tissue-specific loss of function mouse models for CPEB4. The characterization of

these genetic models has revealed that CPEB4-mediated translation is essential for

mammalian organisms to cope with metabolic stress. Moreover, we have identified

the molecular mechanism by which CPEB4 establishes a translational program to

maintain cellular homeostasis during metabolically challenging conditions.

Altogether, our results reveal a novel function of CPEB4 in gene expression

regulation during metabolic stress.

Introduction

INTRODUCTION | 17

1� Gene expression and regulation

The expression of the genomic information is arguably the most important process

to life. It allows the use of the bidimensional information contained in the DNA to

build and maintain exquisite molecular machines with the capacity to survive and

adapt to the environmental challenges. Biochemically, the gene expression pathway

comprises the group of molecular processes that uses the DNA sequence as a

template for the synthesis of specific gene products, commonly proteins. This is

accomplished by the transcription of a sequence of nucleotides from the DNA

into an intermediary molecule called messenger RNA (mRNA), which is then

translated into the sequence of amino acids that forms a protein. This sequential

process of events, known as the central dogma of molecular biology, was already

outlined in 1958 (Crick, 1970; Crick, 1958) and has maintained its validity after

more than 50 years.

Because not all proteins are required to be produced at the same levels and at

the same time, the regulation of the genetic information flow is instrumental

for the control of cell structure and function, and it is the foundation of cell

adaptability to environmental changes. In higher organisms, it becomes even of

greater importance, since the genome contains the instructions to generate and

guide the deployment of all cellular identities during embryonic and postembryonic

development. Beyond that, precise control of genes allows the establishment of

the multiple cell types that will conform the adult organism and will meet the

environmental challenges.

Being such a central process for the maintenance and evolution of life, gene

expression regulation is a complex, interconnected and multilayered process in

which every step of the pathway is tightly monitored and controlled to ensure

18 | INTRODUCTION

optimal cellular adaptation to the environmental and physiological demands

(Moore, 2005) (Orphanides and Reinberg, 2002). However, most of the research

in the last decades has focused on the first step of the pathway: the regulation

of transcription. This has led to the vast accumulation of knowledge regarding

the cis-regulatory elements, trans-acting factors, chromatine structures and

epigenetic modifications that govern the transcription of genes (Kornberg, 1999).

The main reasons of this focus are historical: pioneer studies in the control of

genes were carried out in bacteria (Jacob and Monod, 1961), in which most of

the regulation is takes place at the transcriptional level. However, given that in

eukaryotes transcription and translation are physically separated, evolution had

the opportunity to develop a plethora of additional regulatory steps to temporally

and spatially fine-tune the production of proteins. It was not until the advent of

novel high-throughput techniques that we began to get a glimpse of the post-

transcriptional gene regulatory networks that govern all aspects of the eukaryotic

mRNA life (Lackner and Bähler, 2008) (Halbeisen et al., 2008).

2� Post-transcriptional gene regulation

In eukaryotes, post-transcriptional mRNA processing is composed of several

stages that increase the genome coding capacity and provide additional regulatory

opportunities. Here, we will make a brief overview of the most prominent post-

translational steps.

2�1 Nuclear processing

The post-transcriptional processing and regulation of a gene starts concomitantly

INTRODUCTION | 19

to the transcriptional process (Aguilera, 2005). As soon as the first 20-30 nucleotides

of the nascent pre-mRNA have been produced by the RNA polymerase II, the

mRNA is capped with an m7G structure at the 5’ end (Topisirovic, 2011). The

process occurs in three consecutive reactions: first the mRNA is cleaved, then a

GMP is linked in a reverse orientation by a unique 5’–5’ triphosphate bond to the 5’

end, and finally the GMP is methylated at position N7. The cap structure is necessary

for nuclear export of the mRNA and precludes 5’–3’ exonucleotide degradation

(Topisirovic, 2011). In addition, it is required for proper mRNA translation (see

Translational control of gene expression). These regulatory activities are mediated

by cap-binding proteins: nuclear cap binding proteins complex (nCBC) in the

nucleus and eukaryotic translation initiation factor 4E (eIF4E) in the cytosol.

Most eukaryotic genes contain intronic regions that must be removed to obtain a

mature and functional mRNA (Papasaikas and Valcarcel, 2016). This is generally

achieved by a ribonucleotide complex called the spliceosome, which excise the

introns and ligates together the flanking exons of the pre-mRNA (Papasaikas

and Valcarcel, 2016). This process, far from being constitutive and rigid, allows

the combinatorial removal of different intronic regions expanding the coding

capabilities of a single gene. In addition, it generates broad opportunities to

regulate the production of different protein isoforms from a single locus depending

external and internal signals. The fact that more than 90% of the human genes

are alternatively spliced underscores the importance of this mechanism in the

maintenance of cellular proteostasis (Papasaikas and Valcarcel, 2016).

Once the polymerase has transcribed an entire gene, the pre-mRNA must be

released and its 3’ end must be processed to generate a functional transcript (Di

Giammartino et al., 2011). Except mRNAs encoding for replication-dependent

histones, the 3’ ends of all eukaryotic transcripts are generated by a two-step

reaction: first, the pre-mRNA is endonucleolytically cleaved and released; second,

20 | INTRODUCTION

it is appended with a non-templated polyadenylate [poly(A)] tail. This process

requires the cooperative action of specific regulatory elements in the pre-mRNA

sequence and various RNA binding proteins (RBPs) that will recognize and bind

those elements. The cleavage and polyadenylation specific factor (CPSF) and the

cleavage stimulating factor (CstF) recognize the poly(A) signal (AAUAAA) and

the U/GU-rich element, respectively, on the 3’ region of the pre-mRNA. Then,

the coordinate action of cleavage factors I and II (CFI, CFII), and the subsequent

action of a poly(A) polymerase (PAP), lead to the cleavage and addition of a poly(A)

tail some nucleotides downstream of the poly(A) signal (Di Giammartino et al.,

2011). The coating of the newly synthesize poly(A) by the poly(A) binding protein

nuclear I (PABNI) is essential to increase PAP processivity and to avoid transcript

degradation. Interestingly, the length of the tail varies greatly among species (from

70-80 adenines in yeast up to 250-300 in humans) but it is an essential element for

proper export, stability and translation of mature transcripts in all of them. Until

very recently, it was thought that the 3’ end processing of the transcripts was a

default mechanism. However, genome-wide studies have revealed that, similarly to

the regulation of splicing, more than half of mammalian transcripts can be cleaved

and polyadenylated at various sites, thereby allowing the generation of transcripts

with different 3’ UTRs that will contain or exclude specific regulatory elements that

determine their localization, stability and translation (Elkon et al., 2013). Although

the molecular details of such process are still being worked out, it seems that the

action of specific RBPs acting on regulatory elements of the pre-mRNA is vital for

the 3’ UTR determination (Bava et al., 2013).

Finally, and before reaching the cytoplasm, mature transcripts can be enzymatically

modified on specific nucleotides. Examples of this are RNA pseudouridilation,

cytidine deamination or adenine methylation. In particular, reversible methylation

of adenosines in the position N6 (m6A) has emerged has a widespread modification

of mRNAs and controls their expression and stability (Fu et al., 2014). Considering

INTRODUCTION | 21

that there are more than 100 types of chemical modifications known to occur on the

RNA and that we know very little about the biological function and distribution of

most of them, our knowledge on RNA-epigenetics is expected to rise dramatically

in the near future.

2�2 Cytosolic crossroad

Once the mature mRNA reaches the cytosol, its fate will be predetermine by the

regulatory structures and sequences acquired in the nucleus. As we have seen,

the structure of a mature transcript will be composed of an m7G cap at the 5’

end, followed by a coding sequence that is flanked by two untranslated regions,

and tailored by a poly(A) tail (Figure 1). Every component of the mRNA

exerts specific regulatory tasks that control whether it is translated into proteins,

degraded by exonucleases, repressed in a latent state or localized to specific cellular

compartments.

5’ m7GpppN

5’UTRCAP

AAAAA...A

CODING SEQUENCE 3’UTR poly(A) tail

uORFhairpin miRNAbinding

site

RBPbinding

sites

Figure 1. Eukaryotic mRNA structure and regulatory elements. The m7G cap structure at the 5’

end and the poly(A) tail at the 3’ end are essential components for translation initiation. Secondary

structures (hairpins) and uORFs at the 5’ UTR can negatively regulate mRNA translation. In the

3’ UTR, the presence of miRNA/RBP binding sites can alter mRNA stability and/or translation

through the recruitment of protein complexes that block the cap structure, modify the poly(A) tail

length or cleave the transcript.

22 | INTRODUCTION

3� The cytosolic fate of mRNAs

Here we will analyze the common cytoplasmic fates of an mRNA (namely

translation, degradation, and localization), emphasizing the regulatory function

exerted by its different parts.

3�1 mRNA decay

Once in the cytoplasm, the mRNA has to subsist long enough in order to be

translated. The task is not trivial due to the ubiquitous presence of endo- and exo-

nucleases. The stability of a given transcript is grossly determined by its poly(A)

tail length. Although every transcript is nuclearly endowed with a 250 nt poly(A)

tail, when it reaches the cytoplasm the action of deadenylases can trim the tail

triggering the decay of the mRNA. Transcript deadenylation is commonly followed

by decapping by the decaping enzymes 1 and 2 (DCP1 and DCP2) rendering the

mRNA accessible to 5’-3’ exonucleases and the cytoplasmic exosome which will

degrade it (Halbeisen et al., 2008).

As opposed to this general degradation pathway, the mRNA decay can also be

controlled in a transcript-specific fashion by regulatory sequences located in the

untranslated regions that recruit protein effectors. These effectors can actively

remove the poly(A) tail or trigger the cleavage of the mRNA (see Translational

control by cytoplasmic polyadenylation).

3�2 mRNA localization

RNA localization at specific cellular compartments allows a tight spatial control

INTRODUCTION | 23

of the protein production process. Active transport of mRNAs mediated by RBP-

motor protein complexes is the most widely-studied mechanism of subcellular

RNA localization (Halbeisen et al., 2008). Transcript-specific localization by

this mechanism is achieved by regulatory elements on the mRNA sequence that

recruit RBPs effectors, which repress the translation of the transcript and engage

in transportation to specific cellular compartments, such as the endoplasmic

reticulum, mitochondria, dendrites, etc, (Nagaoka et al., 2012) (Huang et al., 2003).

3�3 mRNA translation

The translational efficiency of an mRNA is fundamentally determined by its

capacity to recruit the ribosomal machinery. Although there are alternative modes

of recruitment (Weingarten-Gabbay et al., 2016), it generally occurs through the

m7G cap-structure (Figure 2).

Cap-mediated translation begins with the formation of the 43S pre-initiation

complex (Gray and Wickens, 1998; Jackson et al., 2010; Sonenberg and Hinnebusch,

2009). This is attained by the recruitment of the ternary complex (TC), formed

by eIF2 (a heterotrymer of α, β and γ subunits), GTP and a methionyl-initiator

tRNA (Met-tRNAiMet) to the 40S ribosomal subunit, which is aided by eIF1, eIF1A

and eIF3. The recruitment of the newly formed pre-initiation complexes to the

transcript is mediated by the eIF4F complex. eIF4F is formed by eIF4E, a cap-

binding protein; eIF4G, a scaffolding protein; and eIF4A, an RNA helicase that

assists the scanning. In addition to eIF4F, the cytosolic poly(A) binding protein

(PABP) is essential to initiate translation. It binds both the poly(A) tail and eIF4G

bridging together the two ends of the mRNA in a process of pseudo-circularization

(Tarun and Sachs, 1995) (Sachs et al., 1997). This conformation, known as the

“closed-loop model”, increases ribosome recycling and has been shown to be

24 | INTRODUCTION

essential for translation initiation, at least in yeast. When the 43S pre-initiation

complex reaches the start codon it forms the 48S initiation complex, which recruits

the 60S subunit to form a 80S translationally competent ribosome.

Figure 2. Cap-mediated translation initiation. Translation begins with the formation of the ternary

complex (TC) by eIF2, GTP and the initiator tRNA. The TC is then recruited to the ribosomal 40S

subunit to form the 43S pre-initiation complex, which is in turn recruited to the mRNA by eIF4F.

The complex then can start scanning the transcript until the initiation codon is reached. At that

point, the 48S initiation complex is formed and the 60S ribosomic subunit is recruited to a form a

80S full ribosome. Modified from Gebauer et al., 2004.

INTRODUCTION | 25

While translation initiation is directly regulated by more than 25 proteins, the

elongation and termination processes require a minimum number of them,

underscoring the preponderance of the initiation step over the other two (Lackner

and Bähler, 2008). However, in recent years, condon usage and acyl-tRNA

availability have emerged as important determinants of translation elongation and

protein synthesis (Richter and Coller, 2015).

4� Translational control of gene expression

Global gene expression analysis in mammalian cells has revealed that translation

efficiency is the single best predictor of protein expression (Schwanhausser et al.,

2011) and is under strong evolutive selection (Khan et al., 2013; Wu et al., 2013),

underlying the importance of this last step in the gene expression cascade.

4�1 Global translational control

The formation of the TC is the major checkpoint that the cell uses to regulate

global protein production (Figure 2). eIF2 is a G-protein that cycles between

and active form (GTP-bound) and inactive form (GDP-bound). eIF2B is the

guanine nucleotide exchange factor (GEF) that regulates the transition towards

the active form by catalyzing the GTP-GDP exchange. However, when eIF2α is

phosphorylated in S51 (in humans), the GEF activity of eIF2B is inhibited such

that the generation of new TC is halted. Several signaling pathways converge on the

phosphorylation of eIF2α by the action of different kinases, what is collectively

known as the integrated stress response (IRS). These kinases are the interferon-

induced double-stranded RNA (dsRNA)-dependent eIF2α kinase (PKR), activated

26 | INTRODUCTION

by viral infection (Galabru et al., 1989); the general control nonderepressible 2

(GCN2), activated by amino acid deprivation (Berlanga et al., 1999); the heme-

regulated inhibitor kinase (HRI), triggered by heme deficiency (Lu et al., 2001),

and the endoplasmic reticulum (ER)-resident kinase (PERK), triggered by the

accumulation of misfolded proteins inside the ER (Shi et al., 1998) (see Translation

at the ER: a surveillance system).

The cap-binding protein eIF4E is another node of general protein synthesis

regulation and is controlled by direct phosphorylation or sequestration by eIF4E

binding proteins (4EBPs) (Richter and Sonenberg, 2005).

Most of the translational activity in a cell occurs at a specialized organelle called

the endoplasmic reticulum (ER), which, given its importance, has develop its own

global translational control mechanism.

4�2 Translation at the ER: A surveillance system

The vast majority of proteins that are destined for cellular organelles, cell membrane

or secretion are synthesized, folded, matured and transported at the ER (Cnop et

al., 2012). Since they account for more than 75% of the proteins produced in a

given cell, their synthesis is one of the most energy consuming cellular processes

(Proud, 2002) and thus, it is tightly controlled. The ER is a specialized eukaryotic

organelle formed by an intricate network of tubes and flattened sacs connected by

a lumen that contains a molecular environment particularly favorable for protein

maturation. In it reside numerous chaperons, protein disulfure isomerases and

peptidyl-prolyl isomerases, which assist in the process of protein folding and

modification. Moreover, the ER lumen chemical properties are also adapted to the

folding process: it contains Ca2+, which is fundamental for chaperone function, 4

orders of magnitude more concentrated than the cytoplasm (0.1mM compared to

INTRODUCTION | 27

10nM). In addition, its redox potential is highly oxidizing thereby benefiting the

formation of disulfide bonds (Oakes and Papa, 2015).

In spite of all the molecular mechanisms set in place to assist it, protein folding is

the most error prone phase of the gene expression cascade. It is estimated that, in

highly synthetic organs like the liver, up to 50% of the proteins fail to assume their

properly folded forms (Hotamisligil, 2010). Moreover, the flux of proteins into

the ER can vary up to one order of magnitude in very short periods of time (Itoh

and Okamoto, 1980). Therefore, cells face the fundamental task of preventing

the accumulation of misfolded proteins in the ER, a condition commonly known

as ER stress (Ron and Walter, 2007) (Walter and Ron, 2012). Far from being an

unlikely event, accumulation of misfolded proteins by a loss of ER homeostasis can

be caused by a variety of events such as increased in protein synthesis, decreased

proteosomal activity, excess or deficiency of nutrients, alterations in calcium

homeostasis or redox hoemostasis, etc., (Wang and Kaufman, 2016).

Considering the wide range of perturbations that can affect ER function, eukaryotic

cells have evolved a surveillance and quality control system to monitor whether the

folding capacity of the ER is in balance with the protein synthesis demands. This

system is composed by a signal transduction pathway called the unfolded protein

response (UPR) (Figure 3) (Ron and Walter, 2007) (Walter and Ron, 2012), which is

conserved from yeast to mammals. The UPR is formed by three different branches

regulated by three sensors: IRE1α, PERK and ATF6α (Bertolotti et al., 2000)

(Harding, 1999). All of them are transmembrane proteins located at the ER and

characterized by a luminal domain with the capacity of detecting the accumulation

of misfolded proteins. In homeostatic conditions, BiP, which is an abundant ER

chaperone, is bound to the intraluminal domains of these sensors and precludes

their activation. However, since BiP presents higher binding affinity to misfolded

proteins, when they accumulate inside the lumen of the ER, BiP is titrated away

28 | INTRODUCTION

from the UPR sensors leading to their activation (Bertolotti et al., 2000) (Shen et

al., 2002) (Figure 3).

Figure 3. The UPR pathway. Three parallel branches comprise the UPR pathway. Every branch

contains a different sensor: ATF6α, IRE1α and PERK. The accumulation of misfolded proteins in

the ER-lumen titrates BiP away from the luminal domain of the three sensors. Once released from

BiP, ATF6α travels to the Golgi were the action of the proteases S1P and S2P release a cytosolic

fragment that travels to the nucleus and activates the transcription o several genes. The release

of BiP from IRE1α and PERK triggers their dimerization and activation. IRE1α catalyzes the

cytosolic splicing of XBP1u, generating XBP1s, which is active in transcription. In addition, IRE1α

activation triggers the degradation of ER-located transcripts (RIDD). Active PERK phosphorylates

eIF2α, which attenuates global protein synthesis and stimulates the translation of uORF-containing

mRNAs such as Atf4. The three pathways cooperate to promote adaptation to the stress conditions

and to increase cell survival. Modified from Wang et al., 2016.

When BiP is released from ATF6α, it unmasks two Golgi localization sequences

INTRODUCTION | 29

(GLS) in its luminal domain that drive its transport to the Golgi (Shen et al., 2002).

Then, the ATF6α cytoplasmic N-terminus, that contains a basic leucine zipper

motif that functions as a transcription factor, is proteolitically cleaved allowing its

translocation to the nucleus where it binds to cAMP response elements (CRE) and

ER stress response elements (ERSE) of the DNA promoting the production of

proteins that mediate protein folding, ER-associated degradation (ERAD) and ER

biogenesis (Figure 3).

Conversely, IRE1α and PERK are activated by the dimerization-driven trans-

autophosphorylation that occurs when BiP dissociates from their luminal

domains. In the case of IRE1α, autophosphorylation activates the cytoplasmic

endoribonuclease catalytic domain, which carries out the splicing of the Xbp1

mRNA by an unconventional spliceosomal-independent mechanism (Figure 3)

(Yoshida et al., 2001) (Calfon et al., 2002) (Lee et al., 2003). The mature XBP1

mRNA generates a protein able to bind and stimulate the transcription of genes

under the control of ERSE. On the other hand, activated PERK phosphorylates

eIF2α by its cytoplasmic kinase domain thereby eliciting a global suppression of

translation (Wek et al., 2006) (Harding, 1999) (see The UPR translational regulation

is mediated by uORFs). Interestingly, eIF2α phosphorylation leads to the specific

production of the transcription factor ATF4 and other genes. ATF4 activates the

transcription of genes that allow adaptation to the ER stress conditions (Figure

3).

In summary, the three pathways act together in order to achieve the same goal:

increase the cell folding capacity by expanding the ER and elevating the amount

of chaperons while reducing general protein synthesis to decrease the number of

client proteins in the ER (Walter and Ron, 2012). However, this adaptive response

cannot last forever. If the ER stress is excessive, the adaptive response switches

into a cell-death response (Rutkowski et al., 2006). This is achieved by the ATF4-

30 | INTRODUCTION

mediated production of the transcription factor CHOP, which then triggers the

core mitochondrial apoptotic cascade suppressing the prosurvival factor BCL2,

while stimulating proapoptotic BIM, PUMA and NOXA (Tabas and Ron, 2011).

Although the molecular details are still poorly understood, it seems that the duration

and intensity of the insult to the ER determines the life/death switch of the UPR.

4�3 The UPR translational regulation is mediated by uORFs

While in yeast the genetic regulation triggered by the UPR is mainly transcriptional,

as we have seen, in higher eukaryotes it is accompanied by a powerful translational

regulation (Figure 3) (Pavitt and Ron, 2012). The RNAase activity of IRE1a

has been shown to degrade several ER-located mRNAs in a process named

regulated IRE1-depdendent decay (RIDD) (Figure 3) (Hollien et al., 2009). More

interestingly, it has also been shown to regulate microRNA levels during ER stress

to specifically modulate gene expression (Chitnis et al., 2013).

However, the UPR branch that more strongly regulates mRNA translation during

the UPR is the PERK pathway (Harding et al., 2000). It simultaneously regulates

bulk protein synthesis and specific protein production. As described above, PERK

inhibits global protein synthesis by phosphorylating eIF2α, thereby hindering the

production of new TCs. (Figure 3). By doing so, it shrinks the workload placed

on the ER allowing it to recover before protein synthesis ensues.

Paradoxically, there are a number of mRNAs whose translation benefit from eIF2α

phosphorylation. They all contained specific regulatory sequences in their 5’ UTR

called upstream open reading frames (uORFs) (Harding et al., 2000) (Harding et

al., 2003). These sequences repress the translation of the main ORF in homeostatic

conditions while promoting its translation whenever the TC availability is reduced.

This regulatory system has been thoroughly studied on Atf4 mRNA (Vattem and

INTRODUCTION | 31

Wek, 2004) (Figure 4). The 5’ UTR of Atf4 mRNA contains two uORFs, uORF1

is only three codons long and uORF2 overlaps with the main ORF. When the

ribosome scans from the cap towards the main ORF and encounters uORF1, it

gets translated. After termination, the ribosome reinitiates the scanning from that

point and, given the high abundance of TCs, it is highly probable that the 40S

subunit is able to form the 43S pre-initiation complex before reaching the uORF2.

Thus, the translation of the uORF2 prevents the decodification of the main ORF

(Figure 4). On the contrary, in conditions of low TC availability caused by eIF2α

phosphorylation, most of the 40S subunits exiting the uORF1 will not acquire a

TC before reaching the uORF2, therefore bypassing its inhibitory function and

allowing the translation of the main ORF (Figure 4). This mechanism applies to

all mRNAs regulated by uORF although with small variations depending on the

number of uORFs, their position and whether or not they overlap with the main

ORF. Accordingly, this counterintuitive but effective system has been shown to

allow the production of some of the main UPR mediators such as ATF4, ATF5,

CHOP and GADD34 (Pavitt and Ron, 2012).

Figure 4. uORF-mediated translational regulation. During normal conditions (high TC abundance)

the two uORFs of the Atf4 mRNA get translated (in blue and red) precluding the translation of the

main ORF (green). However, upon eIF2α phosphorylation, the low availability of the TC hinders

the formation of the 43S pre-initiation complex before reaching the uORF2. In this conditions a

32 | INTRODUCTION

higher number of scanning subunits reach de main ORF allowing it to be translated. Modified from

Pavitt and Ron, 2012.

Altogether, the combination of transcriptional, translational and post-translational

regulation entailed by the UPR generates exceptionally rapid and sharp changes in

the cellular proteome to recover ER homeostasis.

4.4 mRNA-specific translational control: The RBPs

The translational regulatory mechanisms presented so far, based on eIF2α

phosphorylation and eIF4E inhibition, targeted the recruitment of the ribosomes

to the transcripts in a generalized manner. Conversely, there are a group of

regulatory machineries that act in a transcript-specific manner. They rely on the

presence of cis-acting sequences in the mRNAs that recruit specific trans-acting

factors: the RNA-binding proteins (RBPs).

RBPs are central components in RNA metabolism. They regulate all aspects in the

life cycle of the mRNA: RNA synthesis, processing, maturation, export, stability,

transport and translation (Gerstberger et al., 2014) (Singh et al., 2015) (Glisovic et

al., 2008). Indeed, as soon as the pre-mRNA is produced it is rapidly packed into

ribonucleoprotein (RNP) complexes through the binding of a myriad of RBP to its

different elements. The identity and dynamics of single RBPs on mRNP complex

varies through the life cyle of the mRNA through processes of remodelling (Singh

et al., 2015). Intriguingly, RBPs not only regulate every step of RNA metabolism

but also connect them together generating regulatory networks with the capacity

to counteract perturbations in one step of the pathway by modulating others

(Orphanides and Reinberg, 2002). Moreover, it has been propose that RBPs form

RNA-regulons, in which every RBP controls several mRNAs encoding proteins

INTRODUCTION | 33

that function in the same biological process (Keene, 2007). For instance, the neuro

oncologic ventral antigen (Nova) RBP coordinates the translation of a subset of

mRNAs involved in the inhibitory activity of synapses in neurons (Keene, 2007).

Recent advances in high-throughput sequencing and mass spectrometry have

revealed that in eukaryotes up to 20% of total expressed protein-coding transcripts

encoded RBPs (Gerstberger et al., 2014), underscoring that RNA metabolism is not

only and essential cellular process but also one that demands the highest protein

copy number. Furthermore, RBPs are commonly subjected to post-translational

modifications allowing them to integrate signals and to couple the expression

of genes to the cellular demands. Interestingly, RBPs are highly enriched in

intrinsically disorder domains or low-complexity sequences that enables them to

constitute dynamic aggregates like P-bodies, stress-granules or transport granules

(Gerstberger et al., 2014).

While some RBPs bind structural elements of the mRNA such as the m7G cap

(cap-binding proteins, CBPs) or the poly(A) tail (poly(A)-binding proteins, PABPs)

and some are left behind by the splicing machinery (exon-junction complex, EJC),

there are some others that bind with high specificity to regulatory sequences

commonly located in the UTRs. The latter category constitutes a highly interesting

subgroup of RBPs that endow eukaryotic cells with the potential to deploy

particular regulatory features to every single mRNA.

34 | INTRODUCTION

5� Translational control by cytoplasmic polyadenylation

As we have seen above, the poly(A) tail of the mRNA is a dynamic structure that

controls many aspects of mRNA metabolism like transcript stability, translation

and transport. Thus, modulating the poly(A) tail length of mature mRNAs

constitutes a fast and reversible regulatory mechanisms than enables the cell to

quickly respond to environmental stimuli by simultaneously degrading, inhibiting

or activating the translation of different subsets of transcripts in a temporal and

spatial regulated manner (Weill et al., 2012).

This regulation is accomplished by the coordinate action of cytosolic deadenylases,

microRNA-RBPs (miRNPs) and cytoplasmic polyadenylation complexes. All

these RNP-complexes recognize cis-regulatory elements either by direct protein

interaction or through small RNA molecules (Weill et al., 2012). Most commonly,

the target sequences are located in the 3’ UTR of the mRNAs, the only region of

the transcript that is not cleared by the progression of the ribosome. Given the

pseudo-circularized conformation that mRNAs acquire, the RBPs bound to the 3’

UTR are in close proximity to both the poly(A) tail and the cap-structure, the two

main determinants of mRNA stability and translation.

5.1 The CPEB-family of proteins

One interesting family of RBPs is the Cytoplasmic Polyadenylation Element

Binding Protein (CPEB)-family, which regulate mRNA translation and stability

through the modulation of the poly(A) tail length of target mRNAs (Ivshina et al.,

2014) (Fernandez-Miranda and Mendez, 2012).

INTRODUCTION | 35

In mammals, the CPEB-family is comprised of four paralogs: CPEB1-4. Because

CPEB2-4 contain higher sequence identity compared to CPEB1, they are usually

clustered into a separated subfamily (Fernandez-Miranda and Mendez, 2012)

(Ivshina et al., 2014) (Figure 5). Comparison between species reveals that the

CPEBs are better conserved across species than across paralogs (Wang and

Cooper, 2010) suggesting strong evolutionary selective pressure to maintain their

different identities. However, some species contain different number of CPEBs,

for instance, in the arthropod Drosophila melanogaster there are two CPEBs called

Orb1 and Orb2, whereas in the mollusc Aplysia californica there is just one, aCPEB

(Fernandez-Miranda and Mendez, 2012) (Figure 5).

Spisula solidissima (clam) cCPEB

CPEB4 CPEB3

CPEB2

CPEB1

C. elegans Cpb-2

C. elegans Cpb-3

Drosophila Orb

Aplysia californica aCPEB

Zebrafish, Zorba

Xenopus, xCPEB1

Human, hCPEB1

Mouse, mCPEB1

Zebrafish, zCPEB2

Xenopus, xCPEB2

Human, hCPEB2Mouse, mCPEB2

C. elegans Cpb-1

Drosophila Orb2

Zebrafish, zCPEB3

Zebrafish, zCPEB4Xenopus, xCPEB3

Xenopus, xCPEB4

Human, hCPEB3

Human, hCPEB4

Mouse, mCPEB3

Mouse, mCPEB4

C. elegans Fog-1 (Cpb-4)

0.1

CPEB2-4

36 | INTRODUCTION

Figure 5. Phylogenetic analysis of the CPEBs. (Upper panel) Unrooted phylogenetic tree of the

main orthologs and paralogs in the CPEB-family. In red, CPEB1 orthologs; more distant, and

in blue, green and yellow, the orthologs of CPEB2, CPEB3 and CPEB4, respectively. (Lower

pannel) Structural comparison of the CPEBs from various species. The colour of the RRMs (RNA

recognition motives) reflects their differential similarities. Poly Q refers to polyglutamine stretches

and the double vertical red bars depict the ZZ domains. Modified from Fernández-Miranda and

Méndez, 2012; and from Ivshina et al., 2014.

It is now well established that, although with different affinities, all the CPEBs are

able to bind a specific type of AU-rich sequence, called cytoplasmic polyadenylation

element (CPE) that resides in the 3’ UTR of some mRNAs, indicating that their

target population of mRNAs may overlap (Igea and Mendez, 2010) (Novoa et

al., 2010; Ortiz-Zapater et al., 2012). The first CPE motives identified consisted

on UUUUAAU or UUUUAU sequences, although several non-consensus variants

have been subsequently discovered. Alternatively, and based on in vitro selex

experiments, it was proposed that CPEB3-4 may recognize a U-rich loop motif

(Huang et al., 2006); however, it remains to be tested whether this interaction

occurs in any biological context. Genome-wide studies have suggested that up

to 20% of the genome is under the control of the CPEBs (Belloc and Mendez,

INTRODUCTION | 37

2008; Novoa et al., 2010; Pavlopoulos et al., 2011) indicating that CPEB-mediated

translational control might be a widespread mechanism to fin-tune cellular protein

production in metazoans.

Structurally, the CPEBs share common features. They all contain an unstructured

N-terminal regulatory domain (NTD) and a RNA-binding C-terminal domain

(CTD) (Figure 5 lower panel). While the former is highly variable and is subject

to various post-transcriptional modifications depending on the CPEB (Mendez et

al., 2000a; Mendez et al., 2000b; Drisaldi et al., 2015; Theis et al., 2003), the latter

is very much conserved, suggesting that while all the CPEBs can bind the same

targets, they may integrate different signals and affect the mRNA targets in various

ways, generating a complex regulatory network that couples the expression of the

CPE-containing mRNAs to several signalling pathways.

The RNA-binding properties of the CTD are conferred by the presence of two

RNA-recognition motives (RRMs) and a ZZ-domain, in which the latter adopts

a cross-braced zinc coordination topology (Merkel et al., 2013). The crystal

structure of the domain was recently solved and revealed that target specificity

is mostly given by the first RRM. Initially, it was thought that the ZZ domain

mediated the recruitment of sumoylated proteins (Merkel et al., 2013), however,

and unexpectedly, the structure of the whole CTD revealed that the ZZ domain

is involved in the RNA binding activity, although it does not confer specificity

(Huang et al., 2006) (Hake et al., 1998).

The sequence similarity degree of the NTD between the different CPEBs is quite

low, ranging from 40% among the CPEB2-4 subfamily to 10% when compared

to CPEB1. Hence, every CPEB-NTD is modified by different enzymes. CPEB1

is known to be phosphorylated by Aurora A, CDK1 and PLK1, and it undergoes

rapid degradation mediated by a PEST-box domain (Mendez et al., 2000a) (Mendez

et al., 2000b). Phosphorylation, although by different kinases, also seems to regulate

38 | INTRODUCTION

CPEB4 function: the hyperphosphorylation of its NTD regulates its activity by

modulating its aggregation status (data not published). Alternatively, SUMOylation,

monoubiquitination and polyQ-sequence-dependent-aggregation seem to be the

regulatory modifications that control CPEB3 activity, at least in neurons (Drisaldi et

al., 2015) (Theis et al., 2003); on the other hand, the posttranslational modifications

that regulate the activity of CPEB2 remain to be elucidated.

Several mechanisms have been shown to control the expression levels of the CPEBs.

In the case of CPEB1 and CPEB4, they have been described to autoregulate their

own expression through binding to the CPE-elements harboured in their own 3’

UTRs (Igea and Mendez, 2010) (Hu et al., 2014). In the case of CPEB4, this has

been shown to generate both positive and negative regulatory feedback loops that

confer ultra-sensitivity properties to CPEB4 expression. Cpeb1 gene undergoes

epigenetic silencing during gastric cancer progression by heavy methylation of its

promoter. A recent report on circadian transcriptomics in mouse liver revealed

the rhythmic expression of Cpeb2 and Cpeb4 mRNAs during the day (Kojima et

al., 2012). This study showed that the circadian expression of the CPEBs in liver

correlates with the differential polyadenylation of various transcripts at different

times of the day, suggesting a strong temporal regulation of the CPEBs’ function

by the molecular clock (Kojima et al., 2012).

Regarding the regulatory activities exerted by the CPEBs on their target mRNAs,

they have been mainly studied during Xenopus laevis oocyte maturation. However, in

recent years, data unveiling their functions and mechanisms of action in different

biological contexts has started to emerge.

INTRODUCTION | 39

5.2 The activatory and repressory functions of the CPEBs

CPEB1 has been shown to exert both activatory and repressory activities on the

translation of target mRNAs depending on its phosphorylation status (Fernandez-

Miranda and Mendez, 2012). When unphosphorylated, CPEB1 represses the

translation of the mRNAs by a mechanism that is still controversial. There are three

alternative and mutually exclusive models proposed for CPEB1-repressing function

(Figure 6, left): (1) CPEB1 inhibits translation by recruiting the deadenylase

PARN to the mRNA, which outcompetes the cytosolic poly(A)-polymerase (Gld2)

and therefore shortens the poly(A) tail and precludes the circularization of the

transcript (Barnard et al., 2004); (2) CPEB1 inhibits ribosomal recruitment to

the transcript by recruiting Masking, an eIF4E binding protein that hinders the

formation of the eIF4F complex (Stebbins-Boaz et al., 1999); (3) CPEB1 recruits

another eIF4E binding protein, eIF4E-T, with the same outcome as in model (2)

(Minshall et al., 2007).

Figure 6. Regulatory functions of CPEB1. The CPE-element and the hexanucleotide are

40 | INTRODUCTION

represented as a rectangle and an hexagon in the 3’ UTR on the mRNAs, respectively. Three models

have been proposed for CPEB1-mediated mRNA repression (left). Progesterone treatment leads

to the phosphorylation of CPEB1 (depicted as a red star), which triggers the recruitment of the

polyadenylation complex (right). The increase in length of the poly(A) tail allows the binding of the

PABP and the formation of the “close loop”.

On the other hand, the CPEB1 activatory mechanism seems to be better established:

when CPEB1 gets phosphorylated by Aurora A, the repression complex gets

remodeled by an change in the affinity of CPEB1 towards CPSF (Mendez et al.,

2000b), which ejects PARN from the complex and allows Gld2 to extend the length

of the poly(A) tail (Kim and Richter, 2006) (Figure 6, right).

An important element that determines the activity of CPEB1 is the arrangement of

the CPE-elements within the 3’ UTR (Pique et al., 2008). Two CPEs separated by

an optimal distance of 12 nucleotides seem to be necessary for CPEB1-mediated

repression, most probably because they favor the formation of a dimer (Pique et

al., 2008). On the contrary, translational activation only requires the presence of

a single CPE, within a range of 100 nucleotides from the poly(A) signal, being 25

nucleotides the optimal distance predicted. Quite recently it has been proposed

that transient dimerization of CPEB1 might also regulate its function (Lin et al.,

2012).

Beyond meiosis, the mechanism of action of CPEB1 is quite conserved, at least

in neurons (Udagawa et al., 2012), although with important variations. CPEB1

represses translation by recruiting another eIF4 binding-protein called Masking,

and, when phosphorylated by Aurora A or CamKII kinases, it stimulates translation

through the recruitment of the cytoplasmic poly(A) polymerases Gld2 or Gld4

(Atkins et al., 2004).

CPEB2 has been proposed to repress transcript translation by inhibiting the

INTRODUCTION | 41

elongation phase through the interaction with eEF2 (Chen and Huang, 2012).

Although the molecular details are not clear, CPEB2 has also been described to

activate the translation of some target mRNAs (Hagele et al., 2009).

Most of the translational activity of CPEB3 has been studied in the context of

synapse activity. There, it has been found that CPEB3 can either drive mRNA

degradation through actively promoting deadenylation or activate translation

when is subjected to monoubiquitination by Neurilized1 (Pavlopoulos et al.,

2011), cleaved by Calpain2 or when it forms aggregates with prion-like properties

(Drisaldi et al., 2015) (Fioriti et al., 2015; Stephan et al., 2015). More recently, it has

been proposed that CPEB3 might be regulated by PKA- and CamKII-mediated

phosphorylation in the NTD (Kaczmarczyk et al., 2016).

Finally, CPEB4 has been proposed to act as a repressor of translation during

terminal erythroid differentiation through its interaction with eIF3 (Hu et al.,

2014). However, the rest of studies that analyzed CPEB4 mechanism of action

revealed and activatory role during oocyte maturation, somatic cell cycle and tumor

progression (Igea and Mendez, 2010) (Novoa et al., 2010) (Ortiz-Zapater et al.,

2012), mediated by the recruitment of the canonical cytoplasmic polyadenylation

machinery.

5.3 The CPEBs in somatic tissues – CPEB4

Given that CPEB-mediated regulation is instrumental for meiotic progression and

that CPE-regulated mRNAs are highly abundant in all biological processes, the

study of the CPEBs in non-germinal tissues has dramatically increased in recent

years revealing important function for each CPEB beyond meiosis.

Several reports have unveiled that CPEB1-mediated translation regulates mitotic

42 | INTRODUCTION

cell cycle (Novoa et al., 2010), cellular senescence (Burns and Richter, 2008), tumour

development (Nagaoka et al., 2015) (Burns and Richter, 2008), inflammation

(Ivshina et al., 2015) and synaptic plasticity (Udagawa et al., 2012). Moreover,

CPEB1 seems to regulate mammalian organism homeostasis by repressing the

translation of PTEN and STAT3 mRNAs in liver, thereby modulating the hepatic

insulin-signalling pathway (Alexandrov et al., 2012). Therefore, CPEB1 seems to

be important for both pathogenic and non-pathogenic genetic regulation.

Although the somatic functions of CPEB2 remain to be discovered, CPEB3 plays

an essential role in the regulation of hippocampal memory formation (Fioriti et al.,

2015), synaptic plasticity (Drisaldi et al., 2015) and thermosensation (Fong et al.,

2016).

CPEB4-mediated regulation is important for the mitotic progression of tumoral

cells (Novoa et al., 2010), embryonic erithroid differentiation (Hu et al., 2014),

cell survival (Chang and Huang, 2014) (Kan et al., 2010), tumour malignancy

(Ortiz-Zapater et al., 2012) and pathological angiogenesis (Calderone et al., 2016)

(Garcia-Pras et al., 2016). In fact, CPEB4 KO mice partially suffer from neonatal

lethality (Hu et al., 2014), although a similiar model does not (Tsai et al., 2013),

underscoring the importance of CPEB4-mediated regulation. Interestingly, CPEB4

protein localizes at the endoplasmic reticulum of neurons, yet the significance

of this localization remains unknown (Kan et al., 2010). CPEB4 seems to play a

protumoral role in cancer progression (Ortiz-Zapater et al., 2012) (Boustani et al.,

2016) (Sun et al., 2015) (Chen et al., 2015) (Zhong et al., 2015) (Hu et al., 2015)

although some reports point in the opposite direction (Peng et al., 2014).

In spite of the growing literature regarding the roles of CPEB4, it remains to

be shown whether CPEB4-mediated regulation plays any role in adult tissue

homeostasis.

INTRODUCTION | 43

6� Cellular metabolism and translation

6.1 Bidirectional link between the UPR and cellular metabolism

Given that the ER is the major anabolic organelle in a cell, it plays a central role in

cellular metabolism. It regulates protein synthesis, the most energetically expensive

cellular process (Proud, 2002). It also constitutes a metabolic hub where many of

the enzymes involved in glucose and lipid metabolism are located (Fu et al., 2012).

Consequently, it is not surprising that the ER-surveillance system and the cellular

metabolism are closely interconnected to couple the cellular metabolic needs to

the ER folding capacity, and vice versa. Accordingly, researchers have found strong

functional connections between the ER the cellular bioenergetic powerhouse, the

mitochondria (Malhotra and Kaufman, 2011).

The ER and the mitochondria are functionally and physically interconnected, with

20% of the mitochondria surface in close contact with the ER membrane through

specialize contact sites called mitochondrial associated membranes (MAMs). These

interactions have been shown to be fundamental for Ca2+ signaling, lipid transport,

energy metabolism and cell death (Malhotra and Kaufman, 2011). Importantly,

upon ER stress the UPR increases the production of ROS and the liberation of

Ca2+ from the ER reservoir into de cytoplasm. This Ca2+ is uptaken by the MAMs

and modulates the activity of mitochondrial dehydrogenases and nitric oxide

synthases (Malhotra and Kaufman, 2011). Through these mechanisms, the UPR

pathway couples the ER folding status with the mitochondrial energy production

activity. Thus, prolonged UPR activity hinders mitochondrial function. Further

evidences support that in conditions of irreversible ER stress the UPR triggers

mitochondrial-dependent apoptosis through the action of CHOP and the release

of Ca2+ (Wang and Kaufman, 2016). Conversely, and although the molecular details

44 | INTRODUCTION

are not clear, there are evidences that mitochondrial dysfunction can lead to ER

stress. Therefore these two organelles seem to control and regulate one another in

order to maintain protein folding, energy homeostasis and Ca2+ equilibrium inside

the cell.

6�2 The UPR and the etiology of metabolic disease

Taking in consideration the strong conection between the ER and cellular

metabolism, it is no surprise that the UPR is involved in the pathogeneses of

several metabolic diseases such as diabetes type 2 and non-alcoholic fatty liver

disease (NAFLD) (Fu et al., 2012) (Malhi and Kaufman, 2011) (Gentile et al., 2011)

(Rutkowski et al., 2008) (Oakes and Papa, 2015). NAFLD is a condition in which

the hepatic-accumulated fat accounts for more than 5% of the total liver weight, in

the absence of chronic alcohol consumption or any other liver disease (R. and AJ.,

2013). NAFLD has become the most common cause of liver disease worldwide

with a prevalence of around 30% in the western world, and up to 75% in obese

populations (R. and AJ., 2013) (Feldstein, 2010).

In mammals, the liver acts as a metabolic hub that process and distributes all

nutrients to the different organs. Under homeostatic conditions, the main sources

of fat that arrive at the liver are plasma non-sterified fatty acids (FFA) released from

the white adipose tissue or absorbed by the intestine from dietary sources (Figure

7). Moreover, the liver actively synthesizes lipids by a process known as de novo

lipogenesis. Once inside the liver, FFAs can enter the mitochondrial β-oxidation

pathway for energy production or can be esterified into triacylglicerides (TGA).

TGA are then stored inside the hepatocytes as lipid droplets or secreted into the

bloodstream as very low density lipoproteins (VLDL). Therefore, excessive fat

accumulation inside the liver can occur as result of increased FFA uptake, increased

INTRODUCTION | 45

lipogenesis, decreased β-oxidation or decreased lipoprotein secretion (Figure 7)

(Postic and Girard, 2008) (Browning and Horton, 2004) (Anderson and Borlak,

2008).

Fatty acids

Triglycerides

Lipoprotein(VLDL

Adipose tissue

Glucose

Intestine

mitochondrialβ-oxidation

endoplasmic reticulum

Figure 7. Hepatic lipid metabolism. Hepatic free fatty-acids (FFA) come from dietary sources,

adipose tissue release or de novo lipogenesis. FFA can be then used as a source of energy through

mitochondrial β-oxidation. Alternatively, they can be esterified into TGA molecules that are stored

in the form of cytosolic lipid droplets or secreted as very low-density lipoproteins (VLDLs).

Hepatic steatosis occurs when the influx of FFA increases (uptake or synthesis) or when the output

pathways capacity decreases (β-oxidation or lipoprotein secretion).

In some individuals, hepatosteatosis progresses to steatohepatitis, in which the

accumulation of fat is accompanied by liver fibrosis and inflammation (Angulo,

2002). NAFLD constitutes a major risk factor for the development of cirrhosis,

liver failure, cardiovascular disease and hepatocellular carcinoma (Angulo, 2002).

46 | INTRODUCTION

However, the molecular mechanisms underlying the disease are poorly understood

(Feldstein, 2010). For instance, there is no current explanation for the variable

susceptibility to the disease among different human populations, the increasing

prevalence in individuals with low BMI or why the disease progresses at different

speeds depending on the individual.

In recent decades the UPR has been proposed to play a central role in the

progression of NAFLD (Ozcan et al., 2004). On the one hand, hyperlipidemia

causes hepatic ER stress through a variety of mechanism: an imbalance in the

cellular lipid pool that leads to alterations of membrane fluidity, increased in

protein palmitoylation and generation of ROS, all of which are triggers of the

UPR (Wang and Kaufman, 2014) (Volmer and Ron, 2015). Conversely, several

genetic mouse models have demonstrated that the UPR pathway is key to limit

hepatic lipid accumulation. Genetic ablation of any of the UPR branches leads to

the development of hepatosteatosis in mice by decreasing the secretion of liver

lipoproteins (by inducing the degradation of ApoB100), decreasing mitochondrial

β-oxidation of fatty acids and/or increasing hepatic lipogenesis (Lee et al., 2008)

(Wang, 2012) (So, 2012).

Indeed, chemical chaperons are being tested for the treatment of NAFLD

(Maerkeen, 2010) (Ozcan et al., 2006). These compounds prevent ER stress by

stabilizing protein-folding intermediates. In a seminal work, researches proved

that treating obese mice with 4-phenyl butyric acid (PBA) or taurine-conjugated

ursodeoxycholic acid (TUDCA) alleviated ER stress and prevented the development

of hepatosteatosis and insulin resistance (Ozcan et al., 2006). Indeed, TUDCA

treatment has been shown to ameliorate hepatic and muscular ER stress and to

increase insulin sensitivity in obese men and women (Maerkeen, 2010). This and

other studies position the UPR/ER stress pathway as an attractive therapeutic

target for drugs against metabolic diseases.

INTRODUCTION | 47

7� The circadian clock

Over the past few years a growing body of evidence has unveiled that almost

every molecular process that occurs in living organisms is under the control of an

endogenous time-keeping mechanism known as the circadian clock (Wijnen and

Young, 2006) (Asher and Schibler, 2011). Molecularly, this system consists on a

network of transcription factors interconnected by positive and negative feedback

loops that generate waves of gene expression with a periodicity of approximately

24 hours (Figure 8). This system can run autonomously, in the absence of

external stimuli, but can also be entrained by environmental cues. In mammals, the

circadian clock is organized in a central pacemaker formed by the neurons of the

suprachiasmatic nucleus (SCN) that maintain proper phase alignment of peripheral

tissue clocks present in all cells (Bass and Takahashi, 2010).

Figure 8. The mammalian circadian clock

architecture. The core of the molecular clock ,

which is located in almost every cell, comprises

positive and negative feedback circuits that

generate autonomous oscillations in gene

expression with a 24 h periodicity. Clock

rhythmicity can be entrained by environmental

and organic cues (input factors) to maintain

cellular synchronicity. Circadian regulation of the

clock-controlled genes (ccgs) allows the temporal

modulation of the physiology, behaviour and

metabolism of mammalian organisms. Modified

from Wijnen and Young, 2006.

48 | INTRODUCTION

7�1 Post-transcriptional gene regulation by the clock

Importantly, the molecular clock modulates organism physiology and behavior

by controlling the expression of hundreds of genes (Eckel-Mahan et al., 2012)

(Figure 8). It is not surprising that most of the studies on the genetic regulation

exerted by the molecular clock focused at the transcriptional level; however, recent

findings have broadened this view by revealing extensive clock-mediated post-

transcriptional circadian gene regulation. Therefore it is now well established that

the processes of splicing, alternative polyadenylation, translation and cytoplasmic

polyadenylation are regulated by the circdaian clock (Gachon and Bonnefont, 2010)

(Douris et al., 2011) (Janich et al., 2015) (Kojima et al., 2015) (Garbarino-Pico

and Green, 2007) Accordingly, in mouse liver 50% of the rhythmic proteins are

translated from non-rhythmic mRNAs (Atger et al., 2015) and circadian rhythms

exist in anucleated cells (JS. and Reddy, 2011) underscoring the importance of

post-transcriptional mechanism in the cell rhythmic physiology.

7�2 Circadian regulation of mammalian metabolism

As expected, cellular metabolism seems to be under the strict control of the circadian

clock (Bass and Takahashi, 2010) (Asher and Schibler, 2011) (Eckel-Mahan et al.,

2012). In the metabolic organs of mammals, the timing-system plays a central

role in the adaptation and anticipation of the feeding-fasting cycles that constitute

the daily behavior. In accordance, genetic disruption of the clock in mice leads to

a plethora of metabolic alterations such as obesity, diabetes and NAFLD (Birky

and Bray, 2014) (Bray and Young, 2011) (Wijnen and Young, 2006) (Reinke and

Asher, 2016). Interestingly, clock disruption in mice leads to a hepatic steatosis that

can be reverted by the treatment with the chaperone TUDCA, indicating a strong

connection between the circadian clock, ER stress and metabolic disease (Cretenet

INTRODUCTION | 49

et al., 2010). In accordance, intervention strategies targeted at circadian oscillator

components are being tested to prevent or treat NAFLD (Cho et al., 2012) (Reinke

and Asher, 2016) (Solt et al, 2012; Reinke et al, 2016).

At a subcellular level, it is well established that the function of the major metabolic

organelles are under circadian control. The circadian peacemaker generates daily

oscillations of mitochondrial oxidative activity through rhythmic regulation of

NAD+ levels (Peek et al., 2013). Similarly, the UPR signaling pathway shows daily

oscillations in activity (Cretenet et al., 2010).

Although our understanding of this temporal aspect of gene regulation has

tremendously expanded, much remains to be explored about the post-transcriptional

regulatory circuits that coordinate the circadian clock with the cellular metabolism.

Objectives

OBJECTIVES | 51

The goal of this study is to gain a deeper understanding of the post-translational

regulatory mechanism mediated by CPEB4 and their contribution to mammalian

physiology, with special emphasis in cellular metabolism and its circadian control.

The specific goals that we have pursued are:

1. Generation and phenotypic characterization of genetically engineered loss of

function mouse models for CPEB4.

2. Study of the pathophysiological conditions caused by the genetic ablation of

CPEB4 and the underlying etiology.

3. Analysis of the molecular mechanisms that drive the biological alterations

caused by CPEB4 depletion.

4. Identification of CPEB4 mRNA targets and in depth study of the regulatory

mechanism exerted by CPEB4.

5. Analysis of the subcellular localization of CPEB4 in mammalian cells.

6. Characterization of the regulatory mechanisms that control CPEB4 expression.

7. Exploration of the regulation exerted by the circadian clock on the biological

functions of CPEB4.

8. Investigation of the potential impact of these findings in the understanding

and treatment of human disease.

Methods

METHODS | 53

Generation of constitutive and liver-specific Cpeb4 knockout mice

To generate a CPEB4 conditional knockout mouse (CPEB4lox/lox), mouse ES cells

carrying a βgeo-cassette [β-galactosidase gene fused to neomycin resistance gene]

in Cpeb4 intron 1 and loxP sites flanking the exon 2 (clone EPD0060_4_E10; Sanger

Institute) were microinjected into developing blastocysts. The resulting positive

chimeric mice (CPEB4+/loxfrt) were crossed with C57BL6/J mice. Subsequently, the

βgeo-cassette was removed by mating with mice expressing the FlpO recombinase

(Tg.pCAG-Flp), thereby generating conditional knockout animals (CPEB4lox/lox).

To obtain an ubiquitous and constitutive depletion of CPEB4, CPEB4lox/lox mice

were crossed with animals expressing the DNA recombinase Cre under the control

of a human cytomegalovirus minimal promoter (B6.C-Tg(CMV-cre)1Cgn/J). The

excision of exon 2 of the Cpeb4 gene leads to a frame shift in the mRNA generating

several new premature stop codons, resulting in mice that are deficient in CPEB4

protein (CPEB4KO). The offspring was maintained in a C57BL/6J-129S mixed

background. Liver-specific CPEB4 knockout mice (CPEB4LKO) were obtained by

crossing CPEB4lox/lox mice with albumin-Cre transgenic animals. The offspring

was backcrossed for five generations onto the C57BL/6J background. Routine

genotyping was performed by PCR.

Animal studies

Animals were maintained under a standard 12-h light-dark cycle, at 21°C, with free

access to food and water. At 6 weeks of age, male mice were randomly assigned to

either an HFD (60% fat, 20% protein, and 20% carbohydrates; D12492, Research

Diets) or a control regular chow diet (7.42% fat, 17.49% protein, and 75.09%

carbohydrate; RM1, SDS). Only male mice between 2 and 5 months of age were

used, unless otherwise stated. For circadian experiments, mice were caged in a

54 | METHODS

reverse light-cycle room. All experimental protocols were approved by the Animal

Ethics Committee at the University of Barcelona.

Intraperitoneal glucose tolerance test and pyruvate tolerance test

The glucose tolerance test (GTT) was performed on mice after an overnight fast.

Blood glucose was measured a time 0 with a Bayer glucometer, followed by i.p.

injection of glucose (2 g/kg body weight). Blood glucose was measured at 15,

30, 60 and 120 min after injection. Plasma insulin levels were measured with an

Ultrasensitive Mouse Insulin Elisa Kit (Cristalchem) at 0, 30, and 60 min after

injection. For the pyruvate tolerance test (PTT), overnight-fasted mice were injected

with pyruvate (2 g/kg body weight) and glucose levels were measured at 30, 60 and

90 min after injection. For the glucagon tolerance test (GTT), overnight-fasted

mice were injected with glucagon (15 μg/kg body weight) and glucose levels were

measured at 15, 30, 60 and 90 min after injection.

Hepatic VLDL production assay

Mice were fasted for 5 h and then intravenously injected with tyloxapol (500 mg/

kg body weight; Sigma). Aliquots of tail vein blood were taken at 60 and 120

min after injection for plasma triglyceride (TGA) determination using the Serum

Triglyceride Determination Kit (Sigma).

Indirect respirometry analysis

The Oxymax-CLAMs indirect calorimetry system was used to determine

metabolic phenotypes. Mice were left to acclimate in individual metabolic cages

METHODS | 55

for 2 consecutive days prior to data collection. Metabolic parameters (oxygen

consumption, carbon dioxide production and locomotor activity) were then

measured every 20 min for 3 consecutive days.

Primary hepatocyte isolation

Collagenase perfusion (0.5 U/mL) via the inferior cava vein was used to isolate

hepatocytes from fed 12-week-old male mice. Cells were suspended in Dulbecco’s

Modified Eagle’s Medium (DMEM) supplemented with 10 mM glucose, 10% fetal

bovine serum (FBS), 100 nM insulin, and 100 nM dexamethasone and then seeded

onto plates treated with 0.01% (w/v) collagen solution (Sigma). After 3 h, cells

were washed with PBS, and the media were replaced with fresh serum-free DMEM

without glucose for 16 h.

Mouse embryonic fibroblast (MEF) isolation

MEFs were extracted from 14.5-day-old embryos, following standard procedures

(Hogan et al. 1994). Briefly, pregnant female mice were euthanized at day 14.5 of

gestation, and embryos were extracted. After removal of heads and livers, embryos

were minced, incubated in trypsin for 5 min, disaggregated, and resuspended in

culture medium. MEFs were cultured in high glucose-DMEM supplemented with

10% FBS, 2 mM L-glutamine and 1% penicillin/streptomycin at 37°C in a 5% CO2

incubator. Only primary MEFs up to passage 5 were used in all the experiments

except for the circadian synchronization analysis in which 3T3-immortalized MEFs

were used.

56 | METHODS

Mouse embryonic fibroblast clock synchronization

WT and CPEB4KO immortalized MEFs were starved in 0.5% FBS-containing

DMEM for 2 consecutive days. Next, their clocks were synchronized with a 100 nM

dexamethasone shock and protein and RNA samples were collected every 6 h for

2 days.

Fatty acid oxidation, uptake and incorporation

Palmitate oxidation to CO2 was measured in primary hepatocytes grown in 6-well

plates. On the day of the assay, cells were washed in Krebs-Ringer bicarbonate

HEPES buffer (KRBH buffer: 135 mM NaCl, 3.6 mM KCl, 0.5 mM NaH2PO4,

0.5 mM MgSO4, 1.5 mM CaCl2, 2 mM NaHCO3, and 10 mM HEPES pH 7.4) that

contained 0.1% BSA. Cells were preincubated at 37ºC for 30 min in KRBH 1%

BSA and then washed again in KRBH 0.1% BSA. Then, they were incubated for 3

h at 37ºC with fresh KRBH containing 2.5 mM glucose and 0.8 mM carnitine plus

0.25 mM palmitate and 1 _Ci/ml [1-14C]palmitate bound to 1% BSA. Oxidation

measurements were performed by trapping the radioactive CO2 in a parafilm-sealed

system. The reaction was stopped by the addition of 40% perchloric acid through

a syringe that pierced the parafilm.

For palmitate incorporation, cells were incubated for 16 h at 37ºC in serum-free

medium containing 0.25 mM palmitate and 1 _Ci/ml [1-14C]palmitate bound

to 1% BSA. On the day of the assay, cells were washed in PBS, and lipids were

extracted by chloroform-methanol extraction. Total lipids were dissolved in 30

μl of chloroform and separated by thin-layer chromatography to measure the

incorporation of labeled fatty acid into phospholipids (PL), diacylglycerol (DAG),

triglycerides (TGA) and non-esterified labelled palmitate (NEPalm).

Glucose production assay

Primary hepatocytes were plated at 1.740.000 cells/60-mm dish, and after 3 h the

medium was replaced by DMEM supplemented with 2 mM glutamine without

glucose, FBS or sodium pyruvate. The day after, cells were treated for 4 h with the

indicated compounds and cell medium was collected for glucose quantification.

Cellular oxygen consumption

Cellular oxygen consumption rate (OCR) was measured using the Seahorse XF24

analyzer and following the manufacturer’s instructions (Seahorse Biosciences).

MEFs were plated at a concentration of 50.000 cells/well, and OCR measurements

were made after the sequential addition of 1 μg/ml oligomycin, 1 μM carbonyl

cyanide p-trifluoromethoxyphenylhydrazone (FCCP), and 1 μM rotenone

plus 1 μM antimycin A (Rot+AA). Oligomycin treatment inhibits complex V,

thus distinguishing the oxygen consumption devoted to ATP synthesis from

the oxygen consumption required to overcome the natural proton leak across

the inner mitochondrial membrane. Applying FCCP uncouples the electron

chain, unveiling the maximal respiratory capacity. Final treatment with rotenone

(complex I inhibitor) plus antimycin A (complex III inhibitor) reveals the non-

mitochondrial respiration. The OCRs measured were normalized to total protein

content, and routine and maximal respiration was calculated by subtracting the non-

mitochondrial respiration to the basal or FCCP-mediated respiration, respectively.

Respiratory chain function analysis from isolated mitochondria

The respiration of isolated liver mitochondria was measured at 37ºC by high-

resolution respirometry with an Oxygraph-2k (Oroboros Instruments, Innsbruck,

58 | METHODS

Austria). Mitochondria were obtained from mouse liver by differential centrifugation

and then resuspended in buffer II (0.5 M sucrose, 50 mM KCl, 5 mM EDTA, 1

mM sodium pyrophosphate, 5 mM MgCl2 pH7.4, and protease inhibitors). 500

μl of mitochondria suspension was brought to a final volume of 2.1 ml with

respiration medium (0.5 mM EGTA, 3 mM MgCl2.6H2O, 20 mM taurine, 10 mM

KH2P04, 20 mM HEPES,1 g/l BSA, 60 mM K-lactobionate, 110 mM sucrose,

pH 7.1) and added to the oxygraph chamber. All respiration measurements were

done in triplicate. Resting respiration (state 2, absence of adenylates) was assessed

by the addition of 10 mM glutamate and 2 mM malate as the complex I substrate

supply, and then state 3 respiration by the addition of 2.5 mM ADP. The integrity

of the outer mitochondrial membrane was established by the addition of 10 μM

cytochrome c. No stimulation of respiration was observed. The addition of 10 mM

succinate provided state 3 respiration with parallel electron input to complexes I

and II. We examined ADP control of coupled respiration and uncoupling control

through the addition of the protonophore carbonylcyanide-4-(trifluoromethoxy)-

phenylhydrazone (FCCP). The addition of 0.5 μM rotenone resulted in inhibition

of complex I, thereby allowing examination of O2 flux with complex II substrate

alone, while 2.5 μM antimycin A was added to inhibit complex III in order to

observe non-mitochondrial respiration. The concentrations of substrates and

inhibitors used were based on prior experiments conducted to optimize the

titration protocols.

Cell culture and analysis

HepG2 and HEK-293 cells were cultured in high glucose DMEM, supplemented

with 10% FBS, 2 mM L-glutamine and 1% penicillin/streptomycin at 37°C and

5% CO2.

METHODS | 59

Treatment of mouse embryonic fibroblasts (MEFs) with tunicamycin (T7765,

Sigma) and thapsigargin (T9033, Sigma) was done as described in (Rutkowski

et al., 2006). Cells were plated at a concentration of 360,000 cells/60-mm dish

and allowed to rest overnight before the application of the stress. Apoptosis was

measured with FITC Annexin V Apoptosis Detection Kit I (BD Pharmigen),

following the manufacturer’s instructions.

Tunicamycin injection and tissue analysis

Mice were injected intraperitoneally with TM at 1 mg/kg body weight. At the

indicated times, mice were killed and livers were extracted. Various lobules were

separated and snap-frozen in liquid nitrogen, formalin-fixed, or frozen in OCT.

Frozen livers were pulverized in liquid nitrogen and stored at -80ºC. Protein

extracts were obtained by resuspending frozen liver powder in RIPA buffer (50

mM Tris, 150 mM NaCl, 0.1% SDS, 0.5% deoxycholate, 1% NP-40, 1 mM EDTA,

1 mM DTT, 1 mM PMSF) containing protease inhibitors cocktail (11873580001,

Roche) and phosphatase inhibitors cocktail 2 and 3 (P5726, Sigma). Samples were

thoroughly vortexed and sonicated for 10 min at maximal intensity, then centrifuged

at 16000 rpm for 10 min at 4ºC. The supernatant was then stored at -80ºC. RNA

was extracted as described below, and lipids were obtained by chloroform-methanol

extraction.

TUDCA treatment

Mice on a high-fat diet (HFD) for 12 weeks received intraperitoneal injections

of 250 mg/kg TUDCA (Merck) twice a day during the last two weeks of HFD

administration. Then, mice were killed and livers were collected as indicated above.

60 | METHODS

Histological analysis

Livers were fixed in 10% neutral buffered formalin and stained with H&E or Sirius

Red. For Oil-red O staining, liver tissue was frozen in OCT, sectioned, and stained.

Immunoblotting

Protein extracts were quantified by DC Protein assay (Bio-Rad), and equal

amounts of proteins were separated by SDS-polyacrylamide gel electrophoresis.

After transfer of the proteins onto a nitrocellulose membrane (GE10600001,

Sigma) for 1 h at 400 mA, membranes were blocked for 1 h in 5% milk, and

specific proteins were labeled with the following antibodies: anti-CPEB4 (Abcam,

Ab83009); anti-ATF4 (Santa Cruz, sc-200); anti-CHOP (Santa Cruz, sc-575); anti-

phosphorylated eIF2α (Cell Signaling, 9721); anti-VDUP1 (anti-TXNIP) (Santa

Cruz, sc-166234); anti-vimentin (Abcam, ab8978); anti-tubulin (Sigma, T9026);

anti-BIP (BD Bioscience, 610978); anti-TIMM44 (BD Biosciences, 612582); anti-

Calnexin (Santa Cruz, 11397); anti-GAPDH (Life Technologies, AM4300); and

anti-Vinculin (Abcam, ab18058).

Immunofluorescence

Cells were fixed with 4% PFA for 10 min followed by permeabilization with 0.1%

Triton X-100 for 5 min. Then, cells were blocked with 10% FBS in 0.03% Triton

X-100 for 1 h. Primary antibodies were incubated overnight at 4ºC followed by 1 h

incubation with the corresponding secondary antibody. Images were obtained on

an inverted Leica TCS SP5 confocal microscopy.

METHODS | 61

RNA analysis

Total RNA was either extracted by TRIzol reagent (Invitrogen) or RNAspin Mini

Kit (GE Healthcare), followed by Turbo DNA-free Kit (Ambion) treatment. One

microgram of RNA was reverse-transcribed with oligodT and random primers using

SuperScript III (ThermoFisher), following the manufacturer’s recommendations.

Quantitative real-time PCR was performed in a LightCycler 480 (Roche) using

SYBRGreen I Master (Roche). The primer sequences are listed in Supplementary

Table. All quantifications were normalized to endogenous control [TATA-binding

protein (TBP) or 18s].

Lipid extraction

Lipids were extracted following the method of Salmon and Flatt. Liver powder

was digested in potassium ethanol overnight, then centrifuged, and diluted in 50%

ethanol. Samples were mixed with 1M magnesium, incubated on ice for 10 min,

and centrifuged. Supernatant was used for TGA quantification.

DNA extraction and mtDNA quantification

Liver powder was digested with Proteinase K at 50ºC for 16 h. DNA was isolated

by standard phenol-chloroform extraction and ethanol precipitation. mtDNA was

quantified by qPCR using primers designed against mitochondrial DNA and a

nuclear encoded mitochondrial gene (SHD).

62 | METHODS

RNA-immunoprecipitation-sequencing analysis

CPEB4KO and wild-type primary hepatocytes were cultured in DMEM

supplemented with 10% FBS, rinsed twice with 10 ml PBS, and incubated with

FBS-free DMEM and 0.5 % formaldehyde for 5 min at room temperature under

constant soft agitation to crosslink RNA-binding proteins to target RNAs. The

crosslinking reaction was quenched by addition of glycine to a final concentration

of 0.25 M for 5 min. Cells were washed twice with 10 ml PBS, lysated with

scraper and RIPA buffer [25 mM Tris-Cl pH7.6, 1% Nonidet P-40, 1% sodium

deoxycholate, 0.1% SDS, 100 mM EDTA, 150 mM NaCl, protease inhibitor

cocktail, RNase inhibitors], and sonicated for 10 min at low intensity with Standard

Bioruptor Diagenode. After centrifugation (10 min, max speed, 4ºC) supernatants

were collected, precleared, and immunoprecipitated (4 h, 4ºC, on rotation) with

10 μg of anti-CPEB4 antibody (Abcam), or rabbit IgG (Sigma) bound to 50 _l of

Dynabeads Protein A (Invitrogen). Beads were washed 4 times with cold RIPA

buffer supplemented with Protease inhibitors, resuspended in 100 _l Proteinase-K

buffer with 70 _g of Proteinase-K (Roche), and incubated 60 min at 65ºC. RNA

was extracted by standard phenol-chloroform. Samples were processed at the IRB

Functional Genomics Facility following standard procedures.

Subcellular fractionation

The endoplasmic reticulum was purified from whole livers of overnight-fasted

mice using the Endoplasmic Reticulum Isolation Kit (ER0100-1KT, Sigma) and

following the manufacturer’s protocol. Then, proteins and RNA were extracted

following the protocols described above.

METHODS | 63

Microscopy analysis

Stained sections were visualized with a Nicon Eclipse Ci-E microscope. Images

from several regions of the tissue sections were then acquired by a Nikon DS-

Fi2 camera. Digitalized images were analyzed by computerized imaging software

Image J.

Plasmid constructions, cell transfections, and luciferase assays

The cDNA fragment encoding the 5’ UTR of CPEB4 mRNA was inserted between

HindIII and NcoI restriction sites in a pGL3-Promoter vector. The resulting

PSV40-CPEB4-Luc plasmid contains the 5’ UTR of CPEB4 fused to a luciferase

reporter downstream of a constitutive SV40 promoter. 0.5 μg of pSV40-CPEB4-

Luc construct were transiently contransfected with 0.5 μg of a Renilla reporter

plasmid pRL-SV40 in HepG2 cells for 24 h using Lipofectamine 3000 (Thermo

Fisher Scientific) following the manufacturer’s instructions. Transfected cells were

then incubated with 0.1 μM TG for 6 h. Luciferase activity was determined with the

Dual-GLO Luciferase Assay System (Promega). RNA was extracted and quantified

by RT-qPCR to normalize for transfection variability.

Statistics

Data are expressed as means±SEM. Dataset statistics were analyzed using the

GraphPad Prism software. Comparisons between groups were carried out with the

Student’s t test and two-way ANOVA followed by the Bonferroni post-hoc test,

and differences under p<0.05 were considered statistically significant (*P<0.05,

**P<0.01, ***P<0.001).

64 | METHODS

Results

66 | RESULTS

1� Generation of Cpeb4 ubiquitous knockout mice

To examine the somatic functions of CPEB4, we generated a global loss-of-

function genetic mouse model (CPEB4KO) by targeted deletion of exon 2 of the

Cpeb4 gene, as described in Methods. Since the excision of exon 2 of the Cpeb4

gene leads to a frame shift that generates several new premature stop codons, these

animals were depleted of CPEB4 protein in all the tissues tested, as demonstrated

by immunobloting (Figure 9a), and immunohistochemistry (Firgure 9b) and had

reduced levels of Cpeb4 (truncated) mRNA (Figure 9c).

CPEB4

Tubulin

+/+ -/-+/-*

WT CPEB4 KO

WT

CPEB4 KO

0.0

0.5

1.0

1.5

CP

EB

4 m

RN

A le

vels

(AU

)

***

a

b

c

Figure 9. CPEB4 depletion by gene-targeting. a, Immunoblot displaying CPEB4 and α-tubulin

protein levels in CPEB4+/+, CPEB4+/- and CPEB4-/- liver extracts. b, Immunohistochemistry

against CPEB4 in WT and CPEB4KO liver sections. Scale bar, 100 μm. c, Cpeb4 mRNA expression

in WT and CPEB4KO livers, normalized to TBP transcript levels.

CPEB4KO mice were born in the expected Mendelian and male/female ratios,

thereby ruling out an essential role of this protein during embryonic development.

However, we observed that not all knockout mice reached the age of weaning

because of a certain degree of neonatal lethality that varied depending on the

RESULTS | 67

genetic background (30% in C57BL6/J, 129/sv mixed background and 94% in

C57BL6/J pure background). A recent report has identified defective embryonic

erytroid differentiation in CPEB4KO livers as the probable cause of this early

neonatal lethality (Hu et al., 2014).

2. CPEB4 prevents the development of fatty liver disease

Comprehensive histological analysis of animals deficient in CPEB4 revealed no

overt phenotype at young ages and under unchallenged conditions. Furthermore,

CPEB4KO mice were viable, fertile and grossly normal compared with wild-type

(WT) mice. Accordingly, CPEB4-deficient mice showed a body weight increase

comparable to that observed in WT animals (Figure 10a). However, when let to

age for 80 weeks, CPEB4KO old mice developed a hepatic steatotic phenotype with

significantly increased liver weight (Figure 10b) and excessive accumulation of

cytosolic lipid droplets (LD) as detected by H&E and Oil Red staining of livers

(Figure 10c), compared with WT aged mice.

To address whether this aging-induced steatotic phenotype can be recapitulated

in young CPEB4KO animals under feeding conditions that favor obesity, we

fed 20 weeks-old CPEB4KO mice with high-fat diet (60% HFD) for 12 weeks.

Indeed, we found that WT mice were able to adapt to a HFD and to remain

nonsteatotic, whereas mice lacking CPEB4 failed to adapt and developed hepatic

steatosis (Figure 11a-c). Thus, HFD-fed CPEB44KO mice also had exacerbated

liver steatosis, characterized by enhanced liver weight (Figure 11a), excessive

intrahepatic triglyceride content (Figure 11b), and substantially increased fat

deposits and accumulation of triglycerides in the cytoplasm of hepatocytes, as

illustrated by H&E and Oil Red histological stainings (Figure 11c). Moreover, and

in agreement with the known inhibitory effect of lipid accumulation in hepatic

insulin signalling (Postic and Girard, 2008), steatotic HFD-fed CPEB4KO mice

68 | RESULTS

Figure 10. CPEB4 depleted mice develop liver steatosis. a, Changes in body weight of WT and

CPEB4KO mice fed with normal diet, n=16-24. b, Liver weight of 80-week-old WT and Cpeb4KO

mice on normal diet normalized to body weight, n=6-7. Scale bar, 100 μm. c, Left: H&E staining

and Oil Red O staining of liver sections from the same mice. Right: Quantification of area occupied

by lipid droplets (LD), showing accumulation of hepatic lipids in Cpeb4KO aged mice.

suffered from fed and fasted hyperglycemia (Figure 11d). In some CPEB4KO

mice, hepatic steatosis was also accompanied by fibrosis, as assessed by Sirius Red

staining (Figure 11e) The intolerance to metabolism imbalance present in CPEB4-

WT

Oil RedH&E

CPEB

4 KO

WT

CPEB4 KO

01234567

Live

r wei

ght (

% o

f b.w

.) *

WT

CPEB4 KO

0

10

20

30

Are

a oc

upie

d by

LD

(% o

f cyt

osol

)

*

c

a b

3 4 5 6 7048

1216202428

Age (weeks)

Wei

ght (

g)

WTCPEB4 KO

RESULTS | 69

Figure 11. HFD accelerates steatosis in CPEB4KO mice. a, Liver weight normalized to body weight

WT

Oil RedH&E

CPEB

4 KO

WT

CPEB

4 KO

a b

c

f g

d e

Fed Fasted0

50

100

150

200

Pla

sma

gluc

ose

(mg/

dl)

WT CPEB4 KO

**

CHOW HFD0

2

4

6

8

Live

r wei

ght (

% o

f b.w

.) ***WT CPEB4 KO

CHOW HFD0

50

100

150

200

Hep

atic

TGA

(mg

g-1 liv

er)

***

WT KO0

20

40

60

Wei

ght (

g)

*

2 3 4 5 6 7 8 9 100

10

20

30

40

50

Weeks on HFD

Wei

ght (

g)

WTCPEB4 KO

* * * * * **

70 | RESULTS

of 20-week-old WT and CPEB4KO mice fed with a regular diet (CHOW) or with a HFD, n=8-15.

b, Triglyceride content of livers from the same animals. c, Representative photograph of the liver

(left), H&E staining (middle) and Oil Red O staining (right) of liver sections from 20-week-old WT

and CPEB4KO mice on HFD. Scale bar, 100 μm. d, Plasma glucose levels of WT and CPEB4KO mice

on HFD in fasted and fed conditions, n>4. e, Sirius Red staining of liver sections from 20-week-old

WT and Cpeb4KO mice on HFD. Scale bar, 100 μm. f, Body weight of 80-week-old WT and KO

mice on normal diet, n=5-8. g, Changes in body weight of WT and CPEB4KO mice fed with a HFD,

n=14-20. *P<0.05; ***P<0.001.

depleted animals was further supported by the increased body weight of aged and

HFD-fed CPEB4KO animals compared to WT counterparts (Figure 11f,g).

Behavioural alterations seemed not to be the cause of the phenotype because we

did not detect any difference in food or water intake (Figure 12a,b) or locomotor

activity (Figure 12c) in the absence of CPEB4. Neither could we find differences

in respiratory exchange ratio (RER) (Figure 12d) or energy expenditure (EE)

(Figure 12e) in CPEB4KO young animals. Taken together, these results indicate

that deficiency of CPEB4 predisposes to liver steatosis in two different settings of

metabolic homeostasis imbalance (i.e., aging and fat-rich diet feeding).

As CPEB1 knockout mice develop hepatic insulin resistance as a result of altered

translation of mediators in the insulin-signalling pathway (Alexandrov et al., 2012),

which is strongly associated with hepatic triglyceride accumulation (Browning and

Horton, 2004), we next analysed whether glucose metabolism defects could be

the cause of the hepatosteatosis in the CPEB4KO mice. CPEB4 deficiency did not

promote by itself any obvious glucose metabolicabnormality, either in vivo in mice

fed a standard diet or in vitro in primary hepatocytes. In this regard, we did not

find any alteration in plasma glucose (Figure 13a), insulin (Figure 13b), or free-

fatty acid levels (Figure 13c), in fed or fasted conditions. Neither were detectable

differences in the plasma glucose or insulin profile during a glucose tolerance test,

RESULTS | 71

Figure 12. Depletion of

CPEB4 does not elicit

alterations in behaviour. a,

Food intake (g/day) during 4

consecutive days of animals

on HFD, n=4. b, Water

intake in a 24h period, n=6.

c-d, 24-hours time course of

locomotor activity (c), RER

(d) and EE (e) of mice fed a

normal diet, n=3.WT KO

0

1

2

3

4

Wat

er in

take

(ml /

day

) n.s.

0.7

0.8

0.9

1.0

1.1

Time of the day

RE

R

WTCPEB4 KO

0

10

20

30

Time of the day

EE

(kca

l/g/h

)

0

200

400

600

Time of the day

Loco

mot

or a

ctivi

ty (A

U)

1 2 3 40

2

4

6

Time (days)

Food

inta

ke (g

/day

)

WTCPEB4 KO

a b

c

d

e

72 | RESULTS

thereby ruling out the direct involvement of CPEB4 in insulin signalling and

contrasting with the role of CPEB1 (Figure 13d,e) Liver gluconeogenesis was

also unaffected in the CPEB4KO animals, as shown in vivo by a glucagon tolerance

test (Figure 13f) or in vitro by a glucose production assay on primary hepatocytes

(Figure 13g). However, in vivo glucose production from pyruvate substrate

measured by a pyruvate tolerance test was enhanced in CPEB4KO mice (Figure

13h) probably indicating the existence of a subjacent alteration of the glucose

metabolism that however did not have a clear repercussion in the ability of the liver

to respond to insulin or to maintain plasma glycemia (Figure 13a,d). Together,

these results suggest that CPEB4 depletion-induced liver steatosis is not due to a

general dysfunction in glucose metabolism.

3. CPEB4 prevents hepatic steatosis autonomously

Given that CPEB4-depleted mice are heavier upon HFD-feeding or aging, a

possible mechanism to account for abnormal hepatic lipid could be increased

lipolysis in visceral and peripheral adipose tissue and subsequent transport of

free fatty acids to the liver (Fabbrini et al., 2010) (Deng et al., 2012). However,

our findings demonstrating absence of behavioral differences (Figure 12a,b,c),

normal levels of circulating free fatty acids in CPEB4KO mice (Figure 13c) and

high CPEB4 protein expression in hepatocytes (Figure 9b), argue against this

possibility. Therefore, we next sought to determine whether hepatic steatosis

arose as a cell-autonomous defect in the knockout hepatocytes. For this purpose,

we mated animals with a floxed Cpeb4 allele with transgenic mice expressing Cre

recombinase under the control of the albumin gene promoter, thus generating

a hepatocyte-specific depletion of CPEB4 in the postnatal state (Postic et al.,

1999). This model is advantageous in two ways: first, the expression of the Cre is

hepatocyte-specific, thereby rules out the contribution of any other hepatic cell

RESULTS | 73

Figure 13. Absence of major alterations in glucose metabolism in CPEB4KO mice. a, Plasma glucose

DMSO FSK cAMP0.00

0.05

0.10

0.15

0.20

Glu

cose

(mg

10-6

cel

ls)

a b c

d e

f g

h

0 30 600.0

0.2

0.4

0.6

Time after glucose injection (min)

Pla

sma

insu

lin (n

g/m

L)WTCPEB4 KO

0 15 30 45 60 75 90105

1200

100

200

300

400

500

Time after glucose injection (min)

Pla

sma

gluc

ose

(mg/

dl) WT

CPEB4 KO

0 15 30 60 900

50

100

150

200

250

Time after glucagon injection (min)

Pla

sma

gluc

ose

(mg/

dl)

WTCPEB4 KO

0 30 60 900

50

100

150

200

250

Time after pyruvate injection (min)

Pla

sma

gluc

ose

(mg/

dl)

WTCPEB4 KO

**

0 6 240

50

100

150

200

250

Time of fasting (hours)

Pla

sma

gluc

ose

(mg/

dl) WT

CPEB4 KO

Fed Fasted0.0

0.5

1.0

1.5

2.0

Pla

sma

insu

lin (n

g/m

L)

Fed Fasted0

0.5

1.0

1.5

2.0

Pla

sma

FFA

(mM

)

74 | RESULTS

levels of fed, 6h- and 24h-fasted mice, n=6-11. b-c, Fed and overnight fasted plasma insulin levels

(b) and free fatty acid (FFA) plasma levels (c), n=8-9. d-e, Glucose levels (d) and insulin levels (e)

during a glucose tolerance test, n=6. f, Glucagon tolerance test after 6h fasting, n=6. g, Glucose

produced by primary hepatocytes in culture after treatment for 4h with vehicle (DMSO), with the

combination of 10 μM forskolin, 20 mM lactate and 2 mM pyruvate (FSK) or with the combination

of 300 μM dibutyryl-cAMP and 100 nM dexamethasone (cAMP), n=3. h, Pyruvate tolerance test

after overnight fasting, n=13. i, Fed and overnight fasted plasma glucose levels of mice fed a HFD,

n>4. *P<0.01; ***P<0.001.

type; second, it is expressed postnataly, so it allows to discard that any alteration

present in the adult animal is originated during the embryonic development. The

resulting CPEB4LKO mice were born at the expected Mendelian ratio and expressed

normal levels of CPEB4 in all organs tested except in the liver, in which it was

completely absent at 8 weeks of age (Figure 14a,b).

Figure 14. Hepatic CPEB4 depletion in liver-specific CPEB4KO. a, Cpeb4 mRNA expression in

WT and liver-specific Cpeb4KO (Cpeb4LKO) livers, n=4. b, Immunoblot displaying CPEB4 and

α-TUBULIN protein levels in different tissues of WT and CPEB4LKO mice.

WT

LiverCPEB4 LKO WT CPEB4 LKO WT CPEB4 LKO

Intestine

CPEB4

Tubulin

Brain

WT LK

O0.0

0.5

1.0

1.5

mC

PE

B4

RN

A le

vels

(AU

)

**

a

b

*

RESULTS | 75

As in global CPEB4KO mice, liver-specific CPEB4 deletion was not deleterious

per se, having little effect on body weight (Figure 15a) and glucose metabolism

(Figure 15b), compared with WT mice. Neither did it increase plasma alanine

transaminase (ALT) levels (Figure 15c), excluding gross hepatic injury. However,

when animals were challenged with a HFD for 12 weeks, CPEB4LKO mice developed

a lipid deposition phenotype, in a manner very similar to global CPEB4KO animals

fed a HFD. Thus, CPEB4LKO exhibited increased liver weight (Figure 15c) and

enhanced intrahepatic triglyceride content (Figure 15d), along with histologically

visible accumulation of triglyceride within hepatocytes (Figure 15e), compared

to WT counterparts. However, CPEB4LKO did not exhibit increased body weight

gain upon HFD administration (Figure 15g), thereby further discarding that

the hepatic steatosis of the ubiquitous KO mice was secondary to the increase

in obesity. These results indicate that the development of fatty liver observed in

CPEB4-deficient mice in response to HFD-feeding relies on a cell-autonomous

defect in hepatocytes, rather than a metabolic alteration in adipose tissue.

4. CPEB4 deletion causes mitochondrial dysfunction and defective lipid metabolism in hepatocytes

To identify the cell-autonomous anomalies that make CPEB4-depleted hepatocytes

prone to abnormal retention of lipids, we next turned our attention to other lipid

metabolism processes of the liver. We found no differences in the expression of

hepatic fatty-acid synthase and stearoyl-CoA desaturase between CPEB4LKO and

WT mice under a control diet (Figure 16a), indicating that the capacity of the

liver to synthesize fatty acids de novo (ie, lipogenesis) was not altered by CPEB4

deficiency and consequently did not contribute to hepatic fat deposition in

CPEB4LKO mice. Neither were there changes in the fatty acid uptake capacity of

primary hepatocytes isolated from CPEB4KO mice, compared with WT hepatocytes

76 | RESULTS

Figure 15. CPEB4 prevents hepatic steatosis autonomously. a, Weight evolution of WT and

WT

Oil RedH&E

CPEB

4 LK

O

WT

LK

O0.0

0.2

0.4

0.6

0.8

1.0

Pla

sma

ALT

(AU

) n.s.

WT KO

CHOW HFD0

2

4

6

8

Live

r wei

ght (

% o

f b.w

.) **

a b

c d

CHOW HFD0

50

100

150

Hep

atic

TGA

(mg

g-1 liv

er)

***

e

f

3 5 7 9 11 13 15 17 19 210

10

20

30

40

Age (weeks)

Wei

ght (

g)

WTCPEB4 LKO

0 20 40 60 80 100 1200

100

200

300

400

500

Time after glucose injection (min)

Pla

sma

gluc

ose

(mg/

dl) WT

CPEB4 LKO

g

2 4 6 8 10 120

20

40

60

Weeks on HFD

Wei

ght (

g)

WTCPEB4 LKO

RESULTS | 77

CPEB4LKO mice fed a regular diet, n=13-15. b, Glucose tolerance test after overnight fasting, n=8.

c, Plasma alanine aminotransferase levels of mice fed a regular diet, n=7-8. d-e, Liver weight (d)

and hepatic trygliceride content (e) of 20-week-old wild type and CPEB4LKO mice fed a HFD, n=9.

f, Representative photograph, H&E staining and Oil Red O staining of liver sections from the same

animals. Scale bar, 100 μm. g, Growth curve of WT and CPEB4LKO mice on HFD, n=11-12.

(Figure 16b). However, we found diminished mitochondrial fatty acid β-oxidation

in CPEB4KO hepatocytes, even after glucagon stimulation (Figure 16c). This effect

was concomitant with augmented incorporation of fatty acids into triglycerides in

CPEB4KO hepatocytes compared to control (Figure 16d).

Accordingly, maximal mitochondrial activity, which is an effective indicator of

the capacity of cells to manage metabolic stress was reduced both in cultured

cells (Figure 16e) and purified mitochondria from knockout livers (Figure 16f)

deficient for CPEB4. This reduction was more evident for complex II activity

than for complex I (Figure 16f), in agreement with the inability of knockout cells

to utilize fatty acids as a source of energy (Figure 16c). Moreover, we did not

find differences in mitochondrial mass, as measured by mitochondrial proteins and

DNA levels (Figure 16g,h). Further evidence of mitochondrial dysfunction was

provided by the observation of reduced plasma levels of ketone bodies (Figure

16i), which are byproducts of fatty acid metabolism in the mitochondria of liver

cells.

We also assessed whether the lack of CPEB4 impaired triglyceride secretion from

the liver via very low-density lipoproteins (VLDL), which could also contribute to

increase hepatic lipid accumulation. CPEB4LKO and WT littermates were injected

with tyloxapol to block plasma lipoprotein hydrolysis and clearance, followed by the

determination of plasma triglyceride (TGA) levels as an index of hepatic VLDL-

TGA secretion. We found that CPEB4LKO mice accumulated TGA in the plasma at

a lower rate than WT mice (Figure 16j), pointing to suboptimal lipoprotein export

78 | RESULTS

Porin

WT CPEB4 KO

TubulinTimm44Vinculin

WT KO0.0

0.2

0.4

0.6

0.8

-Hyd

roxy

butir

ate

(nM

)

*

TAGDAG PL

NEPalm0

4

8

12

16

Est

erific

atio

n (lip

ids

ug

prot

-1)

WTCPEB4 LKO

***

WT KO0.0

0.4

0.8

1.2

Pal

mita

te u

ptak

e (n

mol

min

-1m

g pr

ot-1

)

WT KO0

400

800

1200

mtD

NA

/nD

NA

ratio

FASN SCD10.0

0.5

1.0

1.5

mR

NA

leve

ls (A

U)

n.s. n.s.

c

d e

f

a b

g

h ij

0 1 20

600

1200

1800

Time after injection (hours)

Pla

sma

TG (m

g/dl

)

WTCPEB4 LKO **

WT CPEB4 LKO

Basal +Glucagon0

1

2

3

4

-oxid

atio

n (n

mol

h-1

mg

prot

-1)

****

State3

(CI)

State3

(CI+CII)

Maxim

al

respir

ation

0

100

200

300

400

OC

R (p

mol

s-1

mg-1

) **

WTCPEB4 KO

Routine Maximal0

2

4

6

OC

R (p

mol

min

-1ug

pro

t-1)

*

RESULTS | 79

Figure 16. CPEB4 deletion causes mitochondrial dysfunction and defective lipid metabolism in

hepatocytes. a, FASN and SCD1 gene expression analysis by qRT-PCR of livers from WT or

CPEB4LKO mice, n=8. b, Analysis of palmitate uptake in primary hepatocytes, n=2. c, Analysis

of palmitate β-oxidation in primary hepatocytes from CPEB4KO or WT mice, untreated or

stimulated with 20 nM glucagon, n=2. d, Analysis of palmitate incorporation into triacylglycerol

(TAG), diacylglycerol (DAG), phospholipids (PL) or non-esterified palmitate (NEPalm) in primary

hepatocytes from CPEB4KO or WT mice, n=2. e, Oxygen consumption rate normalized with total

protein content of MEFs using an extracellular flux analyzer in basal (routine) or FCCP-induced

(maximal) respiration, n=2. f, Oxygen consumption rate of isolated mitochondria from mouse liver

analyzed by Oxygraph-2k at State3 (CI), State3 (CI+CII) and State3u (CI+II), n=3. g, Immunoblot

for the indicated mitochondrial markers and loading controls in WT and CPEB4KO liver extracts.

h, mtDNA quantification by qRT-PCR from WT and CPEB4KO mice livers normalized to nuclear

DNA content, n=8. i, Plasma β-hydroxybutirate levels in overnight fasted WT and CPEB4KO

mice fed a HFD, n=7. j, Hepatic VLDL secretion assay in CPEB4LKO and WT animals. Plasma

triglycerides were measured at 1 h and 2 h after tyloxapol intravenous injection; n=12-13. *P<0.05;

**P<0.01; ***P<0.001.

in livers of CPEB4-deficient animals. No differences in body weight were detected

(Figure 15a), thereby precluding the possibility that the observed reduction in

VLDL production was secondary to body weight changes in CPEB4LKO mice.

Taken together, these data suggest that lipid accumulates in the livers of CPEB4-

deficient animals because of a defect in mitochondrial fatty acid oxidation and

respiration, possibly augmented by impaired lipoprotein secretion.

5. CPEB4 regulates the expression of ER-related proteins

Given that CPEB4 is a RNA-binding protein, we reasoned that the connection

between CPEB4 deletion and increased susceptibility to develop fatty livers might be

most readily determined by identifying which mRNAs are bound to, and therefore

are translationally regulated by, CPEB4 in hepatocytes. For this purpose, we

applied RNA-immunoprecipitation-sequencing (RIP-seq) with specific antibodies

against endogenous CPEB4 protein using CPEB4KO hepatocytes and mock IgG

80 | RESULTS

immunoprecipitation as controls (Figure 17a). Illumina sequencing showed that

444 mRNAs were specifically associated with CPEB4 in hepatocytes (logFC>2)

and that their 3’ UTRs were enriched in CPE motifs compared with the whole

transcriptome (Figure 17b). Intriguingly, a significant number of these CPEB4

targets encoded for endoplasmic reticulum-related proteins, such as TXNIP,

HYOU1, DPM3, TAP1, HMOX1, CRELD2 and TOR3a, as revealed by Gene

Set Enrichment Analysis (GSEA) (Figure 17c). Many of these proteins, including

DNAJC2, DNAJC8, ERP44, HYOU1, NPM3, DNAJA1, SACS, PPIL3, PPIB,

PPIF, DNAJA1, SLMAP and TOR3A, are ER-resident molecular chaperones

that participate in protein homeostasis, specifically helping proteins to fold and

preventing their aggregation. Therefore, these data suggest that CPEB4 regulates

the translation of specific mRNAs encoding ER-related proteins.

Figure 17. CPEB4 regulates the expression of ER-related proteins. a, Immunoblot for CPEB4 and

α-TUBULIN from inputs and immunoprecipitated fractions with α-CPEB4 antibody or IgGs in

CPEB4

InputWT KO WT IgG KO

IP

Tubulin

a

b

c

10-610-510-410-310-210-1100

nuclear envelope - ER networkendoplasmic reticulum membrane

organelle lumenintracellular organelle lumenmembrane-enclosed lumenendoplasmic reticulum part

nuclear lumenendomembrane system

cytosolendoplasmic reticulum

p-value adj

GSEA-2

-1

0

1

2

3E

nrich

men

t sco

re

(RIP

vs

trans

crip

tom

e) +CPE

- CPE

RESULTS | 81

WT and CPEB4KO primary hepatocytes. b, Enrichment score of CPE-regulated and -unregulated

transcripts of RIP-targets versus the mouse transcriptome. c, Top enriched cell compartment

categories based on RIP-seq data analyzed by DAVID bioinformatics resources.

6. CPEB4 is located at the ER, but does not localize mRNAs

The specific binding of CPEB4 to ER-related transcripts made it plausible that, in

order to perform its function, CPEB4 would be located at the vicinity of the ER.

Indeed, previous studies in neuronal cultures showed an ER-specific localization

of CPEB4 (Kan et al., 2010). To test whether this localization was also conserved

in hepatocytes, we subjected mouse livers to a subcellular fractionation, which

revealed that CPEB4 was heavily skewed to the ER in livers of WT mice, being

absent in mitochondrial fractions (Figure 18a). Immunostaining of CPEB4 in

primary mouse embryonic fibroblasts (MEFs) also revealed a perinuclear staining

pattern, congruent with the subcellular localization of the ER (Figure 18b).

Figure 18. ER-localization of CPEB4. a, Immunoblot of liver-fractionated extracts for the

indicated proteins. The fractions from total liver (Liver), mitochondria (Mito) and endoplasmic

reticulum (ER) are shown. b, Immunofluorescence of CPEB4 in WT and CPEB4KO MEFs. Nuclei

are stained with DAPI.

WT CPEB4 KO

CPEB4

CPEB4Liver

WT CPEB4 KOMito ER Liver Mito ER

Calnexin

GAPDH

Timm44

a b

82 | RESULTS

Because CPEB1 has been shown to localize mRNAs to specific subcellular

compartments (Weill et al., 2012), we wondered whether an analogous process

could be taking place with CPEB4 at the ER. To specifically examine whether

CPEB4 was necessary to localize specific mRNAs at the ER, we extracted and

sequenced ER-located mRNAs from WT and CPEB4KO livers. Compared with

controls, only 8 mRNAs showed reduced localization to the ER in knockout livers

(Table 2). These mRNAs were probably mislocalized by indirect causes because

none of them was a direct target of CPEB4 according to the RIP-seq data (Table

1). Thus, we concluded that CPEB4 most likely regulates the translation activity,

but not the localization, of specific mRNAs at the ER.

7. CPEB4 deletion sensitizes livers to dietary fat-induced ER stress

Three lines of evidence lead us to hypothesize that CPEB4 might directly participate

in the response to ER-stress: (1) CPEB4 targets are enriched in proteins involved

in ER homeostasis; (2) HFD feeding and aging are known to generate hepatic ER

stress. (3) Hepatic ER stress impairs mitochondrial FFA oxidation and respiration,

together with lipoprotein secretion (Ota et al., 2008; Raabe et al., 1999; Rao and

Reddy, 2001), which are all affected in CPEB4-depleted cells.

To this end, we first determined the intrahepatic expression of a panel of ER stress

markers by real-time qPCR. Following HFD feeding for 12 weeks, the ER stress

markers and downstream inflammatory cytokines upregulation was significantly

exacerbated in livers from CPEB4KO mice (Figure 19), indicating that animals

with CPEB4 deficiency fail to adaptively attenuate ER stress in response to HFD

feeding.

These results suggest that CPEB4 prevents the development of hepatic steatosis

by promoting adaptation to ER stress

RESULTS | 83

Figure 19. HFD-induced ER stress is exacerbated in CPEB4KO mice. a, Gene expression analysis

by qRT-PCR of livers from CPEB4LKO and WT mice fed a normal diet (CHOW) or a high-fat diet

(HFD), n=8.

8. CPEB4 depletion leads to defective adaptation to chronic ER stress

To confirm that the exacerbated liver fat accumulation in the absence of

CPEB4 was a general consequence of activated and unresolved ER stress and

UPR signaling, we next tested the effect of ER stress chemical-inducers in both

CPEB4-deficient animals and cultured cells. Liver-specific CPEB4KO animals were

subjected to a single intraperitoneal injection of tunicamycin (TM), which inhibits

N-linked glycosylation of newly synthesized proteins and subsequently causes

accumulation of unfolded or misfolded proteins in the ER lumen and ER stress.

Forty-eight hours after challenge, CPEB4LKO mice developed profound hepatic

steatosis as evidenced by the presence of livers much lighter in color than WT

counterparts (Figure 20a, left), accumulation of lipid droplets visualized by H&E

staining (Figure 20a, right) and significantly increased liver weight (Figure 20b)

and triglyceride content (Figure 20c), as compared with WT mice treated with

a

WT CHOW

WT HFD

CPEB4 LKO CHOW

CPEB4 LKO HFD

CPEB4 ATF4 XBP1s CHOP TNFa IL60

1

2

3

mR

NA

leve

ls (A

U)

**

*

84 | RESULTS

TM or mice not subjected to stress. These results are consistent with our findings

in HFD-induced ER stress models and suggest that CPEB4 deletion enhances the

susceptibility to both chronic (dietary fat) and acute (TM) ER stress.

Figure 20. CPEB4 depletion leads to defective adaptation to chronic ER stress. a, Representative

photograph and H&E staining of liver sections from WT and CPEB4LKO mice injected with 1

mg TM/kg body weight and killed after 48 h. Scale bar, 100 μm. b-c, Liver weight (b) and hepatic

triglyceride content (c) of the same animals, n=7-8.

Moreover, the transcript encoding the proapoptotic transcription factor CHOP

was dramatically increased in CPEB4LKO mice, compared with WT animals, after

TM-induced ER stress (Figure 21a) and, accordingly, the percentage of apoptotic

cell death was markedly increased in CPEB4KO cells treated with TM for 24h

WT

CP

EB

4 LK

O+ TUNICAMYCIN

Ctl Tunicamycin0

2

4

6

Live

r wei

ght (

% o

f b.w

.)

Ctl Tunicamycin0

50

100

150

Hep

atic

TGA

(mg

g-1 liv

er) **

WT CPEB4 LKO

**

a

cb

RESULTS | 85

(Figure 21b). These results suggest that CPEB4 may be essential to orchestrate

the adaptation to ER stress and, in its absence, the apoptotic branch of UPR is

favoured. Indeed, upon mild ER stress, ATF4 and CHOP upregulation took place

at lower TM doses in CPEB4KO cells than in WT cells (Figure 21c).

Figure 21. Depletion of CPEB4 favours the apoptotic UPR. a, Gene expression analysis by qRT-

PCR of the indicated genes, n=6. b, Apoptotic analysis of WT and CPEB4KO MEFs measured by

flow cytometry as the percentage of annexin V-positive cells after addition of 5 μg/ml TM for 24

h. c, Immunoblot of ATF4, CHOP, BIP and α-tubulin of MEF extracts 24 h after the addition of

the indicated doses of TM.

To buttress the conclusion that the hepatosteatotic phenotype of CPEB4KO mice

was at least in part caused by a failure to cope with ER stress and to attenuate

ATF4

CHOP

0WT CPEB4 KO

TM (ng/ml) 10 25 250 0 10 25 250

BIP

Tubulin

Ctl Tunicamycin0

10

20

30

40

Apo

ptot

ic ce

lls (%

)

WT CPEB4 KO

**

a

WT + TMCPEB4 LKO + TM

CPEB4CHOP

XBP1ATF4

ERDJ4 P58PDI

TNFA IL6FASN

SCD10.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

mR

NA

leve

ls (A

U)

***

b c

86 | RESULTS

the UPR, we treated HFD-fed animals with tauroursodeoxycholic acid (TUDCA),

a chemical chaperon described to protect from the ER stress-effects induced by

HFD (Ozcan et al., 2006). Strikingly, TUDCA treatment restored both liver weight

(Figure 22a) and intrahepatic triglyceride content (Figure 22b) of CPEB4KO

mice to WT levels, completely reverting the effects caused by CPEB4 absence in

the context of HFD-feeding. These results confirm that the hypersensitivity to ER

stress induced by CPEB4 suppression causes the development of liver steatosis in

mice.

Figure 22. TUDCA rescues the CPEB4LKO steatosis. a, Liver weight of WT and CPEB4LKO mice

on HFD for 12 weeks and treated with TUDCA (500 mg/day/mice) for the last 2 weeks, n=6. b,

Triglyceride content of livers from the same animals. *P<0.05; **P<0.01.

9. CPEB4 protein levels are upregulated by the unfolded protein response

To further define the specific UPR step in which CPEB4 may participate, wild-

type and CPEB4KO MEFs were incubated with thapsigargin (TG), which causes

ER stress by inhibiting the sarco/ER Ca2+ pump (Sagara and Inesi, 1991), and the

expression of the UPR markers and CPEB4 were measured at short times (1, 2, 4

and 6 h) after the induction.

HFD HFD + TUDCA

0

50

100

150

200H

epat

ic TG

A (m

gg-1

liver

)

*n.s.

HFD HFD+ TUDCA

0

2

4

6

8

Live

r wei

ght (

% o

f BW

) ** n.s.

a b

RESULTS | 87

We found that CPEB4 protein levels were low under non-stressed conditions in

WT cells, but were potently and rapidly upregulated, in a time-dependent fashion,

by exposure to TG (Figure 23a), with a kinetic similar to that of ATF4, following

eIF2α phosphorylation and preceding CHOP accumulation (Figure 23a). Cpeb4

mRNA levels, as does Atf4 mRNA, moderately increased at early time-points

(Figure 23b) Similar results were obtained by treating the cells with TM (data not

shown).

Interestingly, the absence of CPEB4 in MEFs did not affect the activation of any

of the UPR branches in response to short-term ER stress. Thus, the time courses

of eIF2α phosphorylation, ATF4 and CHOP expression in CPEB4KO cells after

challenge with TG were very similar to those of their WT counterparts (Figure

23a). These observations suggest that CPEB4 synthesis is either downstream of, or

in a parallel pathway to, eIF2α phosphorylation and ATF4 translational activation,

but not upstream.

Because CPEB4 follows the same expression pattern as ATF4, which mRNA is

translationally activated through an uORF-dependent mechanism, we analyzed

Cpeb4 mRNA for the presence of possible uORFs. Indeed, close inspection of the

5’ UTR of Cpeb4 mRNA revealed multiple (8 in rodent and 9 in human) putative

uORFs that were conserved among various mammalian species (Figure 23c).

Ribosome profiling analysis from published data sets (Gao et al., 2015) (GWIPS-

visualization tool) indicated that these regulatory sequences are in fact translated

in non-stressed conditions (Figure 23d) thereby precluding the translation of

the coding sequence. To test whether these Cpeb4 5’ UTR-uORFs could promote,

or capacitate, translation upon the UPR induction by TG, we expressed chimeric

mRNAs with either Cpeb4 5’ UTR or a control 5’ UTR followed by a reporter ORF.

Indeed, Cpeb4 5’ UTR promoted translation of a reporter ORF, when compared

with a control 5’ UTR after TG treatment (Figure 23e) without affecting the

88 | RESULTS

Cpeb4 5’UTR

1 86 72 3 4 5Mouse

1 86 72 3 4 5Rat

1 872 43 5 6 9Human

0Time TG (h) 1 2 4 6 0 1 2 4 6

P-eiF2α

CHOP

CPEB4

Vinculin

ATF4

WT CPEB4 KO

Cpeb4 5’UTR

Control 5’UTR

CPEB4 5’UTR Firefly

5’UTR Renilla

Ctl 5’UTR Firefly

5’UTR Renilla

Ctl

5'UTR

CPEB4

5'UTR

0.0

0.5

1.0

1.5

2.0

Lum

ines

cenc

e (fi

refly

/reni

lla)

***CtlTG

Ctl

5'UTR

CPEB4

5'UTR

0.0

0.5

1.0

1.5

mR

NA

leve

ls (fi

refly

/reni

lla)

n.s. n.s.

a

c

d

b

e

0 1 2 40

1

2

3

Time TG (h)

CP

EB

4 m

RN

A le

vels

RESULTS | 89

Figure 23. CPEB4 levels are regulated by the UPR. a, Immunoblot analysis of WT and CPEB4KO

MEFs treated with 1 μM TG for the indicated proteins. b, Cpeb4 mRNA levels of the same cells,

n=3. c, Schematic of the phylogenetic conservation of Cpeb4 mRNA uORFs among different

mammalian species. d, Visualization of ribosome footprints within Cpeb4 mRNA (in red) and total

RNA-seq reads (in green) by GWIPS-viz tool. The fragment corresponding to the 5’ UTR of the

mRNA is highlighted. e, Left: Scheme of the dual luciferase reporter assay. A Firefly luciferase

reporter under the control of Cpeb4 5’ UTR and a Renilla luciferase reporter control in HepG2 cells

treated with 0.1μM TG for 6h. Middle: Luminescence values before and after the addition of TG.

Right: qRT-PCR expression analysis of the constructs.

stability of the mRNA.

Therefore, these data demonstrate that CPEB4 protein levels are translationally

regulated by the UPR through the uORFs located in its 5’ UTR, and that this

process is probably conserved throughout tissues and mammalian species.

10. CPEB4 is the only CPEB family member induced by the UPR

Because the CPEB family of proteins is composed of 4 members with overlapping

target mRNA populations, we wondered whether the rest of the CPEBs were

under the same UPR-translational regulation or if otherwise this was a specific

mechanism of CPEB4. Bioinformatic comparison of the 5’ UTRs of all 4 Cpebs

showed that Cpeb4 had a disproportionally large 5’UTR sequence compared to

that of the rest of the members (Figure 24a). Furthermore, neither uORFs nor

ribosomal footprints were identified within the 5’ UTRs of Cpeb1-3. Accordingly,

genome-wide ribosome profiling data showed that CPEB4 was the only CPEB

whose translation was stimulated by TG-induced ER stress in MEFs (Reid et

al., 2014) (Figure 24b), in agreement with our results (Figure 23a,b). All these

experiments support the notion that CPEB4 is the only member of the CPEB

90 | RESULTS

family regulated during the UPR, which positions it as a unique regulator of mRNA

translation during ER stress.

Figure 24. CPEB4 is the only CPEB regulated by the UPR. a, Illustration of the mRNA isoforms

from all CPEB-family members in mouse. Arrows indicate the direction of the transcripts and 5’

UTRs are highlighted. b, Total translation of the CPEB-family members assessed by ribosome

profiling in MEFs treated with 1μM TG for the indicated times (analyzed from Reid D.W. et al.,

2014).

11� CPE-regulated mRNAs are activated during the UPR

Mechanistic studies on the CPEB-family of proteins show that they can both

activate or repress the translation of their target mRNAs. As for CPEB4, while

most of the studies show that it acts as a translational activator (Igea and Mendez,

2010; Novoa et al., 2010), others point to a role in repression (Hu et al., 2014).

Therefore, we next sought to determine whether the UPR-dependent increase in

CPEB4 levels results in translational regulation of CPE-containing mRNAs and, if

a

b

0 0.5 1 2 40

4

8

12

Time TG (h)

Tota

l tra

nsla

tion

(AU

)

CPEB1

CPEB2

CPEB3

CPEB4

RESULTS | 91

that is the case, which is the translational activity of CPEB4 in that context. For this

purpose, we used cells expressing a GFP reporter under the regulation of a 3’ UTR

that contains CPE regulatory elements, either functional (CPE) or inactivated by

point mutations (mCPE), and a RFP reporter as normalization control (Giangarra

et al., 2015). None of the reporters contained uORFs in their 5’ UTRs. These cells

were treated with TG and the fluorescence intensity of both proteins was followed

over time. The UPR activation led to a two-fold increase in the production of GFP

over RFP 8 h after the addition of TG (Figure 25a). This activation was both

dependent on the CPE-elements and CPEB4 levels, as it did not take places in the

mCPE reporter or when CPEB4 was depleted (CPEB4 KD) (Figure 25a).

Strikingly, these results indicate that CPEs and CPEB4 promote translation of

specific mRNAs during the UPR, when eIF2α is phosphorylated and general

protein synthesis inhibited, in a similar but independent manner than the uORFs.

Therefore, a CPEB4-driven mechanism may constitute a new branch of UPR-

mediated translation.

To interrogate whether this translational regulation was also occurring in endogenous

CPEB mRNA targets, we analyzed the UPR-induced activation of TXNIP, an ER-

resident protein in charge of maintaining redox homeostasis. TXNIP is induced in

response to ER stress according to ribosome profiling data (Figure 25b) and is one

of the most enriched targets of CPEB4 according to the RIP-seq results (Table

1). Txnip 3’UTR contains several conserved CPEs (Figure 25c). TG treatment of

MEFs led to a marked increased in both TXNIP protein and mRNA peaking at 2 h

after the induction (Figure 25d,e) as previously described (Figure 25b). However,

the absence of CPEB4 completely abolished TXNIP protein induction (Figure

25d), while the mRNA upregulation was maintained (Figure 25e), indicating that

TXNIP is translationally activated by CPEB4 during the UPR.

92 | RESULTS

Figure 25. The UPR activates CPEB4-mediated translation. a, Left: Scheme of the dual fluorescence

reporter assay: a destabilized GPF reporter under the control of a CPE-containing 3’ UTR and a

destabilized RFP reporter control in HEK-293 cells treated with 1μM TG for the indicated times.

Right: fluorescence values for the indicated times. b, Total translation of Txnip assed by ribosome

profiling in MEFs treated with 1μM TG for the indicated times (analyzed from Reid D.W. et al.,

2014). c, Txnip 3’ UTR sequence in different mammalian species. The conserved CPE-elements are

highlighted. d, Immunoblot for the indicated proteins in WT and CPEB4KO MEFs treated with TG.

e, mRNA levels of Txnip quantified by qRT-PCR from the same samples.

d2EGFP CPE CPE

mCPE

CPE

dRFP

d2EGFP mCPE mCPE

dRFP

TXNIP 3’UTR

MouseHuman

Rat

MouseHuman

Rat

MouseHuman

Rat

0

WT CPEB4 KO

1 2 4 6 0 1 2 4 6Time TG (h)

P-eiF2αCHOP

Tubulin

TXNIP

a

b c

d e

CPE

mCPECPEB4 KD

0h 2h 8h0.0

0.5

1.0

1.5

2.0

2.5

Time TG (h)Fl

uore

scen

ce (G

FP/R

FP)

***

*

0 0.5 1 2 40

100

200

300

Time TG (hours)

TXN

IP (t

otal

tran

slatio

n)

0 1 2 4 60

1

2

Hours of TG treatment

Txni

p m

RN

A le

vels

WTCPEB4 KO

RESULTS | 93

Collectively, these data demonstrate that UPR-induced eIF2α phosphorylation,

while inhibiting global protein synthesis, promotes the translation of uORF-

containing transcripts including Cpeb4 mRNA. In turn, CPEB4 activates the

translation of CPE-regulated mRNAs in a uORF-independent manner.

12. CPEB4 mRNA levels are regulated by the molecular circadian clock

Recent studies revealed that the UPR signalling is a rhythmic process in mouse

liver and it is regulated by the peripheral cell-autonomous hepatic circadian clock

(Cretenet et al., 2010). Given that Cpeb4 is known to be rhythmically transcribed in

this organ (Kojima et al., 2012), we decided to investigate the potential impact of

the circadian clock on the hepatic CPEB4-branch of the UPR.

Close inspection of the promoter sequence of the Cpeb4 gene revealed several

E-boxes and CT-rich regions (Figure 26a), which are known binding sites of

the CLOCK-BMAL1 complex, the core component of the molecular clock.

Although these data suggest that CPEB4 gene might be directly regulated by the

hepatic clock, we set out to gain a deeper understanding on the dynamics of Cpeb4

mRNA throughout the circadian cycle, and based on published liver circadian

transcriptomic data sets (Vollmers et al., 2009), we plotted the hepatic Cpeb4 mRNA

levels throughout the circadian cycle over a 24h period in a 12-h ligh-dark cycle.

We found that hepatic Cpeb4 mRNA levels oscillate in a circadian fashion in mouse

liver peaking in the morning (Figure 26b). Interestingly, at this time of the day is

when the hepatic protein secretion rhythms reach their maximum (Mauvoisin et

al., 2014). Both fed or fasted mice maintained Cpeb4 oscillations, indicating that

Cpeb4 rhythmicity does not respond to the feeding status of the animal. However,

the rhythms of Cpeb4 mRNA did not persist in CRY1/CRY2 KO mice, which do

not have functional circadian clocks (Figure 26b). Thus, liver Cpeb4 mRNA levels

94 | RESULTS

Figure 26. Cpeb4 mRNA is regulated by the molecular clock. a, Cpeb4 promoter sequence. E-boxes

and CT-rich regions are highlighted in blue and in green, respectively. b, Cpeb4 mRNA levels from

livers of WT fed or fasted mice, and CRY1/CRY2 double knockout mice at the indicated ZT.

appear to be controlled directly by the liver circadian clock independently of

feeding-related signals.

As one of the target mRNAs identified by RIP-seq was Cry1, a core component

of the molecular clock, we hypothesiezed that CPEB4, apart from regulating ER

function, could be controlling the clock rhythmicity. For this purpose, we analysed

the circadian rhythmicity of livers and MEFs depleted of CPEB4. Gene expression

analysis of the circadian clock genes Per2 and Bmal1 at various times of the day

did not reveal any significant difference in the period or amplitude of the clock

b

a

0 3 5 7 9 11 13 15 17 19 21 23 250

5000

10000

15000

Zeitgeber time (hours)

CP

EB

4 m

RN

A le

vels FED

FASTEDCRY1/CRY2 DKO

Cpeb4 promoter

E-box CT-rich region

RESULTS | 95

oscillations in CPEB4 mutant livers or cells (Figure 27a,b). Therefore, although

Cpeb4 levels are regulated in a circadian manner, Cpeb4 does not appear to be an

intrinsic component of the clock.

Figure 27. CPEB4 depletion does not affect the rhythmicity of the molecular clock. a, Gene

expression analysis by qRT-PCR of Bmal1 and Per2 in WT and CPEB4KO mouse livers at the

indicated ZT. b, Gene expression analysis of the same genes in WT and CPEB4KO MEFs after

clock synchronization by a dexamethasone shock, n=3.

Given that Cpeb4 mRNA is translationally repressed by the uORFs in homeostatic

conditions, we reasoned that the fluctuations originated by the clock in Cpeb4

mRNA could be dampened by the uORF translational repression and that, only

upon ER stress, CPEB4 protein levels could gain circadian rhythmicity. Indeed, we

a

b

0 4 8 12 16 20 240.0

0.5

1.0

1.5

2.0

2.5

ZT (h)

Per

2 m

RN

A le

vels

(AU

) WTCPEB4 KO

0 4 8 12 16 20 240.0

0.5

1.0

1.5

2.0

2.5

ZT (h)

Bm

al1

mR

NA

leve

ls (A

U)

0 1 6 12 18 24 30 36 42 480

1

2

3

4

Per

2 m

RN

A (F

old

chan

ge)

Time (hours after induction)

WT

CPEB4 KO

0 1 6 12 18 24 30 36 42 480

1

2

3

Time (hours after induction)

Bm

al1

mR

NA

(Fol

d ch

ange

)

96 | RESULTS

found that the circadian Cpeb4 mRNA oscillations in the livers of our mice (Figure

28a) were not reflected at the protein level (Figure 28b). However, upon TM-

induced ER-stress CPEB4 protein levels increased in a circadian fashion (Figure

28c). Thus, we detected greater CPEB4 protein induction by TM at the time of

the day when Cpeb4 mRNA levels peaked (ZT2), compared to when it reached its

trough (ZT14). These oscillations were not the consequence of a differential UPR

induction as phospho-eIF2α and ATF4 levels were comparable.

Figure 28. The circadian clock regulates the UPR-mediated CPEB4 activation. a, Cpeb4 mRNA

P-eiF2α

α-actin

ATF4

CPEB4Ctl

TunicamycinZT0 ZT12

*

*

a

b

c

CPEB40

PER2

GAPDH

3 6 9 12 15 18 21 24 ZT (h)

*

Ctl

TM ZT0

TM ZT120

1

2

3

4

5

CP

EB

4 pr

otein le

vels

(normal

ized to

GA

PD

H)

0 3 6 9 12 15 18 210

1

2

3

4

Zeitgeber time (hours)

CP

EB

4 m

RN

A le

vels

RESULTS | 97

levels in fed WT mice liver extracts at the indicated ZTs. b, Immunoblot for the indicated proteins

from the same livers. c, Left: Immunoblot for the indicated proteins in liver extracts from mice

treated with 1μg TM/g of animal and killed 4 hours after. The ZT of the injections are indicated.

Right: quantification of normalized band intensities.

These data suggest that Cpeb4 mRNA oscillation confers circadian sensitivity to

the CPEB4-branch of the UPR response. Therefore, the circadian clock increases

the “CPEB4-potential” activity during the morning, the time of the day when

hepatocytes are under maximal secretory demand. This mechanism presumably

allows hepatic cells to anticipate the ER stress caused by periods of great synthetic

demand.

13. CPEB1/CPEB4 double knockout mice develop hepatosteatosis

According to these and previously published data, CPEB1 and CPEB4 regulate

different aspects of the hepatic metabolism in such way that depletion of CPEB1

leads to hepatic insulin resistance (Alexandrov et al., 2012) while CPEB4 absence

results in hepatic steatosis upon metabolic stress. Because insulin resistance can

indirectly promote hepatic steatosis through increased lipogenesis and fatty acid

availability, a similar condition that the one generated by HFD feeding, we next

sought to investigate the phenotypic consequences of losing both regulatory

activities at the same time. Thus, we mated CPEB4+/- with CPEB1+/- mice in order

to obtain CPEB1/CPEB4 double knockout mice (CPEB1/4DKO). Crosses between

double heterozygous mice (CPEB1+/- CPEB5+/-) gave rise to all six different

genotypes expected. However, only 30% of the expected double-knockout mice

survived the two firs weeks of age. Strikingly, all the surviving double-knockout

mice presented sings of aberrant lipid accumulation in their livers as early as two

months of age as observed by H&E staining of livers (Figure 31a) even when fed

98 | RESULTS

a normal diet.

These data suggest that in a context of CPEB4 absence, the insulin resistance

condition generated by the depletion of CPEB1 accelerates and worsens the

development of hepatic steatosis.

Figure 29. CPEB1/4 DKO mice develop hepatic steatosis. a, Representative H&E staining of liver

sections from WT and CPEB1/4DKO mice 8 weeks old fed a normal diet.

WT CPEB1/4 DKOa

Discussion

100 | DISCUSSION

1. A novel regulatory branch of the UPR is orchestrated by CPEB4

During aging or metabolic stress, such as HFD, liver cells undergo ER stress

giving rise to UPR signalling in order to maintain tissue homeostasis (Hummasti

and Hotamisligil, 2010) (Brown and Naidoo, 2012). The UPR activation leads to

phosphorylation of eIF2α, thereby attenuating general protein synthesis, limiting

the protein load and helping cells to adapt to ER stress (Wang and Kaufman,

2016). Paradoxically, phosphorylated eIF2α selectively increases translation of

mRNAs that harbour uORFs in their 5’ UTR, such as Atf4, Atf5 and Chop mRNAs.

This translational mechanism allows the deployment of a gene expression program

aimed to resolve the accumulation of unfolded proteins by decreasing the number

of client proteins at the ER while simultaneously increasing the amount of a group

of proteins involved in the folding, maturation and removal of misfolded proteins

(Kaufman, 2002). In this study, we unveiled a novel adaptive translational response

to ER stress mediated by the RNA-binding protein CPEB4. This new arm of the

UPR is distinct and complementary to previously described branches, and it is

essential to maintain cellular homeostasis and survival in conditions of ER stress

(Figures 20 and 21). Thus, upon ER stress, CPEB4 mediates the translational

activation of mRNAs that encode a group of proteins enriched in chaperones and

other proteins implicated in ER homeostasis (Figure 17) (Table 1).

Thus, we found that, upon UPR activation and eIF2α phosphorylation not only

uORF-containing mRNAs are transnationally stimulated, but also CPE-regulated

transcripts are activated (Figure 25). Indeed, the CPE-mediated translational

activation during stress is mechanistically independent of the presence of uORFs,

as reporter mRNAs containing CPE-elements but not uORFs get activated upon

TG treatment. Interestingly, we also found that the two mechanisms are highly

intertwined, as CPEB4, the direct mediator of the CPE-mediated activation, is

itself a uORF-regulated mRNA. Moreover, both regulatory elements (uORFs and

DISCUSSION | 101

CPEs) coexist in various mRNAs, such as Txnip, Hyou1 and Cpeb4. We consider

that it will be worth investigating the extent of the regulatory interaction between

the CPEs and the uORFs, and how they synergize to fine-tune the production of

proteins during ER stress.

2. CPEB4 drives the activation of CPE-containing transcripts

In line with previous reports (Igea and Mendez, 2010) (Ortiz-Zapater et al., 2012)

(Novoa et al., 2010), we find that CPEB4 mediates the translational activation

of target mRNAs, even its own (Figures 23a and 25a) (Table 1). However, a

recent study described a repressive function of CPEB4 on mRNA translation

during erythroid differentiation (Hu et al., 2014). It is important to note that in

the latter, to unveil the molecular mechanism of CPEB4 regulation, the authors

used exogenously overexpressed CPEB4 protein way above the physiological

levels. In our experience, the overexpression of CPEB4 or any other CPEB leads

to experimental artifacts such as protein aggregation and cell death. Since these

conditions make the interpretation of the experiments very complicated, we

consider that there could be alternative mechanistic explanations for the molecular

basis of CPEB4 function during erythroid differentiation. Therefore, we consider

that in mammalian cells, and contrary to CPEB1, CPEB4 most likely only activates

translation of target mRNAs.

But, why is CPEB4-mediated cytoplasmic polyadenylation so crucial during

stressful conditions? It is known that poly(A) tail lengthening increases the capacity

of an mRNA to recruit the ribosomal machinery, thereby increasing its translation

efficiency. However, in homeostatic conditions, subtle changes in the length of

the poly(A) tail may only have a modest impact on the translation efficiency of

the mRNA due to the great abundance of the translation initiation complexes

which are available even for the low efficient transcripts. On the contrary, during

102 | DISCUSSION

ER stress, the big drop in TC availability makes the number of transcripts per

initiation complex to increase dramatically. These new scenario generates

enormous competition among the mRNAs for the translation resources thereby

greatly benefiting the mRNAs which are highly efficient in the recruitment of the

ribosome. Under these circumstances, changes in the poly(A) tail length, and thus

in the translation efficiency, may really make the difference between been translated

or not.

The CPEBs, and in particular CPEB4, have been shown to be essential during

the meiotic and mitotic cell cycle. Interestingly, a recent report demostrated that

during the mitotic phase of the cell cycle cap-mediated translation is inhibited

(Pyronnet, 2001), therefore generating a similar translational conditions as ER

stress. According to this model, we predict that CPEB-mediated regulation would

become an essential regulatory mechanism under any condition that increases

transcript competition for the translational aparatus.

3. The strong connection between CPEB4 and the cellular secretory system

Although previous studies already hinted at the potential link between CPEB4

and ER function, the relevance and mechanistic details of this connection have

remained unexplored. The ER-specific location of CPEB4 was originally observed

in cultured neurons (Kan et al., 2010). Intriguingly, upon TG treatment, the authors

observed a quick relocalization of CPEB4 from the ER to the nucleus. Not only do

we not observe this effect in any of our experimental systems, but also we suspect

that the nuclear localization of CPEB4 might be provoked by the extremely high

doses of TG used in those experiments (10μM, one or two orders of magnitude

higher than the standard) that probably lead to an extreme stress response followed

by an abrupt cell death; alternatively, we also think that given the highly specialized

nature of neurons it is plausible that their regulatory mechanisms differ from other

DISCUSSION | 103

cell types.

In another study, researchers reported the direct interaction between CPEB4

and BiP in a high-throughput screening in astrocytic tumor cells (Chen et al.,

2015). Since BiP is exclusively localized inside the lumen of the ER and mRNAs

cannot cross the ER membrane, we find the interpretation of these results fairly

complicated. Further analysis would be needed in order to validate this interaction

and to explore its biological relevance.

Given that we have also demonstrated ER-specific CPEB4 localization in mouse

hepatocytes and probably MEFs (Figure 16a,b), we suspect that this subcellular

localization of CPEB4 might be conserved across cell types. In the same direction,

the only two validated CPEB4 targets described in previous studies are secreted

proteins [VEGFα (Calderone et al., 2016) and TPA (Ortiz-Zapater et al., 2012)

in endothelial and pancreatic tumoral cells, respectively], which are synthesized,

maturated and transported at the ER, further supporting the strong connection

between CPEB4 and the secretory system.

4. The circadian clock modulates the hepatic function of CPEB4

In this study we also found that a second regulatory mechanism impacts and

conditions the UPR-mediated regulation of CPEB4 levels and activity. Thus, we

showed that the translational regulation exerted by the UPR on CPEB4 protein

levels is dependent on the circadian oscillating Cpeb4 mRNA levels in liver (Figure

28), thus leading to changes in the amplitude of the CPEB4 response to ER stress

along the day.

Through the analysis of published genome-wide datasets (Cretenet et al., 2010) we

unveiled that Cpeb4 transcription is directly regulated by the peripheral molecular

clock in liver (Figure 26b). Hepatic circadian gene expression has been shown

104 | DISCUSSION

to be controlled both by the central clock (located in the supachiasmatic nucleus

of the brain) or by the hepatic peripheral clock (located in every liver cell). The

regulation exerted by the central clock on liver transcripts is indirect and mediated

by changes in the feeding behaviour of the animal. Given that CPEB4 oscillations

are maintained both in fed and fasting conditions, but are lost in clock-deficient

mice, reveals that the peripheral clock is the direct mediator of Cpeb4 mRNA

oscillations. Moreover, the presence of CLOCK-binding sites within the Cpeb4

gene promoter (Figure 26a) further support this conclusion.

Although previous reports had already revealed the circadian expression and

translation of Cpeb4 mRNA (Kojima et al., 2012) (Janich et al., 2015), to our

knowledge, this is the first time that the functional relevance of these oscillations is

addressed. Although there are many of biological advantages in synchronizing an

ER-protective mechanism to the daily oscillations in hepatic metabolism, it remains

to be explored why hepatocytes rely more on the CPEB4 system at earlier times of

the day than at night. We speculate that, since the hepatic protein secretion rhythms

peak in the morning (Mauvoisin et al., 2014), precisely when Cpeb4 mRNA level

reaches its maximum, the circadian clock might work to protect the ER integrity

by increasing the CPEB4-mediated ER stress response when protein synthesis

demand is at its highest.

5. The distinctive role of CPEB4 inside the CPEB-family

Although all four members of the CPEB-family bind the same CPE-element (Afroz

et al., 2014), thereby having theoretical overlapping mRNA target populations

(Igea and Mendez, 2010; Novoa et al., 2010), CPEB4 is the only member whose

expression is regulated by ER stress (Figure 24), and the only one acting as a

crucial regulator in UPR. In fact, and to our knowledge, this CPEB4 function in

stress response has no precedent among RNA-binding proteins. On the other hand,

our results demonstrate a functional interaction between CPEB1 and CPEB4 in

DISCUSSION | 105

the regulation of hepatic metabolism (Figure 29). The simultaneous depletion of

CPEB1 and CPEB4 aggravates the phenotypes present in CPEB4KO mice, thereby

indicating that the two CPEBs cooperate, although by different mechanisms, to

maintain proper hepatic metabolism.

From an evolutionary perspective, the diversification of functions between the

CPEB paralogs would allow organisms to increase their genetic regulatory potential.

It is interesting to note that, while evolution has conserved the same RNA-binding

specificity for the different CPEBs, especially among CPEB2-4 (Afroz et al.,

2014), their N-terminal regulatory domain (NTD) has undergone strong sequence

divergence. As a result, while all the CPEBs bind the same mRNA targets, they are

subject to distinct regulatory inputs through diverse post-translational modifications

in their NTD. In addition, not all the CPEBs are simultaneously expressed in all

cell types.

In our view, this differential regulation endows the CPEB-family with the potential

to build up complex regulatory networks that modulate the translation of CPE-

containing mRNAs based on information integrated from various signalling

pathways. Interestingly, the mRNAs encoding the CPEBs harbour CPE-elements

in their 3’ UTRs, thereby generating cross-regulatory interactions that could confer

robustness and flexibility to the network.

6. CPEB4 protects from NAFLD development

This work also underscores that the CPEB4-regulated branch of the UPR could

have substantial implications in the pathogenesis of non-alcoholic fatty liver

disease (NAFLD), one of the most common causes of chronic liver disease that

can progress to non-alcoholic steatohepatitis (NAS), fibrosis, cirrhosis and liver

cancer (Angulo, 2002; Michelotti et al., 2013). Strikingly, in the US, NAFLD is

currently the second leading cause of liver transplantation, and it is expected

106 | DISCUSSION

to rise in the coming decades. In spite of its alarming prevalence and the fatal

progression that it entails, there is currently not a single pharmacological treatment

approved for NAFLD (Angulo, 2002). Therefore, improving our understanding

of the molecular basis of the disease is key for the development of treatments and

diagnostic techniques, which is becoming a global health priority.

In our study, we believe that we have deepen our understanding of the

pathophysiological mechanisms that originate NAFLD by uncovering a new

genetic regulatory circuit that is essential to prevent abnormal hepatic lipid

accumulation in liver. Therefore, we found that during the increase in hepatic ER

stress caused by age or HFD-feeding, CPEB4-mediated translation is required to

sustain mitochondrial respiration and β-oxidation (Figure 16c,e,f) and hepatic

lipoprotein secretion (Figure 16j). In the absence of CPEB4, hepatocytes are

rendered susceptible to metabolic stress thereby accumulating excessive amounts

of lipids.

We think that our results could help in the identification of potential therapeutic

targets for this life-threatening disease. To advance in this direction, further

experiments will be needed to address whether the artificially activation of the

CPEB4-pathway in steatotic livers could serve as a protective mechanism from

uncontrolled ER stress and lipid accumulation in conditions of hyper-nutrition

or advanced age. Unpublished results from our lab have unveiled that CPEB4

activation relies on the hyper-phosphorylation of its N-terminal domain.

Interestingly, we suspect that the UPR not only triggers a rise in CPEB4 levels but

also in its phosphorylation status, as revealed by a shift in the mobility of CPEB4

upon TG treatment (Figure 23a). Investigating the kinase(s) responsible of such

phosphorylation will greatly contribute the development of CPEB4-targeted

therapies.

Contrary to CPEB1 function in insulin signalling, we found that CPEB4 is

DISCUSSION | 107

mostly involved in lipid metabolism. However, we did detect and increase in the

gluconeogenic capacity of CPEB4-depleted livers upon pyruvate administration

(Figure 13h). This contrasts with the normal glucose production observed

upon glucagon stimulation in mice (Figure 13f) or primary hepatocytes (Figure

13g) depleted of CPEB4. We think that these results could reflect a glucose

metabolism defect in CPEB4KO cells that is only revealed upon conditions of

gluconeogenic-substrate saturation. However, since these conditions rarely occur

in any physiological contexts, we do not consider the function of CPEB4 a major

determinant of glucose metabolism.

The relevance of CPEB4 in human disease is further supported by the identification,

through Genome Wide Association Studies (GWAS), of several single nucleotide

polymorphisms (SNPs) within the Cpeb4 gene associated to human pathology.

There are two genetic variants of Cpeb4 (Rs6861681 and Rs1106693) that strongly

correlate with obesity-related traits (Heid et al., 2010) (Comuzzie et al., 2012). And

there are two more (Rs17695092 and Rs17763373) associated to inflammatory

and prion-derived diseases (Jostins et al., 2012) (Mead et al., 2012), two disorders

characterized to be highly dependent on ER stress and the UPR pathway.

Importantly, every SNP identified within the Cpeb4 gene is located in intronic

regions, indicating that these variants might affect the production or the nuclear

proccessing of the Cpeb4 mRNA, thereby altering the levels of CPEB4 protein

or favouring the production of alternative isoforms. It would be interesting to

test whether individuals with the different variants present changes in the CPEB4

protein, which could endow them with different susceptibilities to ER stress.

There is increasing evidence that the UPR is closely associated with cell differentiation

(Reimold AM et al., 2001) (Zhang K1, 2005), stem cell functions (van Galen et al.,

2014) (Mohrin M et al., 2015) (Wang et al., 2014), and cancer development and

vascularization (Clarke et al., 2014), processes for which high CPEB4 levels are

108 | DISCUSSION

also required (Hu et al., 2014) (Garcia-Pras et al., 2016) (Ortiz-Zapater et al., 2012)

(Calderone et al., 2016). In fact, CPEB4 is essential during embryonic terminal

erythroid differentiation (Hu et al., 2014) and its depletion during this process

leads to partial early neonatal lethality. Interestingly, mice knockout of components

of the PERK-signaling axis of the UPR also present perinatal lethality (Scheuner

D et al., 2001). Mice knockout for PERK or non-phophorilable mutants of

eIF2α die very early after birth of hypoglycemia associated with defective hepatic

gluconeogenesis(Scheuner D et al., 2001). It would be very interesting to test

whether ER stress-derived hypoglycemia contributes to the demise of CPEB4-

depleted newborns.

In conclusion, this study has unveiled a novel role of CPEB4 in the UPR pathway

in vivo. This novel CPEB4-driven adaptive response to ER stress adds an important

player in the UPR field and might be critical for understanding the pathogenesis of

NAFLD and other diseases.

Conclusions

110 | CONCLUSIONS

The present study identifies CPEB4 as a novel regulatory component of the UPR

pathway thereby uncovering a novel translational regulation mechanism that is

key for cellular ER stress adaptation. Furthermore, we have found that this gene

regulatory mechanism is biologically relevant and prevents the development of

hepatic steatosis in mice.

The main conclusions reached in this study are the following:

1. CPEB4 ubiquitous knockout mice develop hepatic steatosis upon aging or

high-fat diet feeding.

2. The prevention of hepatic steatosis mediated by CPEB4 is a cell

autonomous mechanism.

3. CPEB4-depleted hepatocytes show impaired mitochondrial respiration,

FFA β-oxidation and VLDL secretion.

4. In liver, CPEB4 preferentially binds transcripts encoding proteins involved

in ER homeostasis. Although it is located at the ER, CPEB4 does not

control mRNA localization.

5. The absence of CPEB4 renders cells vulnerable to HFD- or TM-induced

ER stress.

6. CPEB4 is translationally and transcriptionally induced by the UPR signalling

cascade, and this regulation is not shared by any other CPEB.

7. The 5’ UTR of Cpeb4 mRNA confers translational activity during conditions

of ER stress and eIF2α phosphorylation.

8. ER stress triggers the translational activation of CPE-containing mRNAs

by a mechanism independent of uORFs and mediated by CPEB4.

CONCLUSIONS | 111

9. Txnip mRNA translation is activated by CPEB4 during ER stress.

10. The mRNA levels of CPEB4 are circadianly controlled by the hepatic

molecular clock, which provides a temporal regulation of the UPR-

mediated activation of CPEB4.

11. CPEB1 and CPEB4 cooperate through independent mechanisms to

maintain hepatic metabolism homeostasis.

References

REFERENCES | 113

Afroz, T., Skrisovska, L., Belloc, E., Guillen-Boixet, J., Mendez, R., and Allain,

F.H.T. (2014). A fly trap mechanism provides sequence-specific RNA recognition

by CPEB proteins. Genes & Development 28, 1498-1514.

Aguilera, A. (2005). Cotranscriptional mRNP assembly: from the DNA to the

nuclear pore. Curr Opin Cell Biol 17, 242-250.

Alexandrov, I.M., Ivshina, M., Jung, D.Y., Friedline, R., Ko, H.J., Xu, M.,

O’Sullivan-Murphy, B., Bortell, R., Huang, Y.T., Urano, F., et al. (2012). Cytoplasmic

polyadenylation element binding protein deficiency stimulates PTEN and Stat3

mRNA translation and induces hepatic insulin resistance. PLoS Genet 8, e1002457.

Anderson, N., and Borlak, J. (2008). Molecular mechanisms and therapeutic targets

in steatosis and steatohepatitis. Pharmacol Rev 60, 311-357.

Angulo, P. (2002). Nonalcoholic Fatty Liver Disease. N Engl J Med 346, 1221-1231.

Asher, G., and Schibler, U. (2011). Crosstalk between components of circadian and

metabolic cycles in mammals. Cell Metab 13, 125-137.

Atger, F., Gobet, C., Marquis, J., Martin, E., Wang, J., Weger, B., Lefebvre, G.,

Descombes, P., Naef, F., and Gachon, F. (2015). Circadian and feeding rhythms

differentially affect rhythmic mRNA transcription and translation in mouse liver.

Proc Natl Acad Sci U S A 112, E6579-6588.

Atkins, C.M., Nozaki, N., Shigeri, Y., and Soderling, T.R. (2004). Cytoplasmic

polyadenylation element binding protein-dependent protein synthesis is regulated

by calcium/calmodulin-dependent protein kinase II. J Neurosci 24, 5193-5201.

Barnard, D.C., Ryan, K., Manley, J.L., and Richter, J.D. (2004). Symplekin and

xGLD-2 are required for CPEB-mediated cytoplasmic polyadenylation. Cell 119,

641-651.

114 | REFERENCES

Bass, J., and Takahashi, J.S. (2010). Circadian integration of metabolism and

energetics. Science 330, 1349-1354.

Bava, F.A., Eliscovich, C., Ferreira, P.G., Minana, B., Ben-Dov, C., Guigo, R.,

Valcarcel, J., and Mendez, R. (2013). CPEB1 coordinates alternative 3’-UTR

formation with translational regulation. Nature 495, 121-125.

Belloc, E., and Mendez, R. (2008). A deadenylation negative feedback mechanism

governs meiotic metaphase arrest. Nature 452, 1017-1021.

Berlanga, J.J., Santoyo, J., and De Haro, C. (1999). Characterization of a mammalian

homolog of the GCN2 eukaryotic initiation factor 2alpha kinase. Eur J Biochem

265, 754-762.

Bertolotti, A., Zhang, Y., Hendershot, L., Harding, H., and Ron, D. (2000). Dynamic

interaction of BiP and ER stress transducers in the unfolded-protein response. Nat

Cell Biol 2, 326-332.

Birky, T.L., and Bray, M.S. (2014). Understanding circadian gene function: animal

models of tissue-specific circadian disruption. IUBMB Life 66, 34-41.

Boustani, M.R., Mehrabi, F., Yahaghi, E., Khoshnood, R.J., Shahmohammadi,

M., Darian, E.K., and Goudarzi, P.K. (2016). Somatic CPEB4 and CPEB1 genes

mutations spectrum on the prognostic predictive accuracy in patients with high-

grade glioma and their clinical significance. J Neurol Sci 363, 80-83.

Bray, M.S., and Young, M.E. (2011). Regulation of fatty acid metabolism by cell

autonomous circadian clocks: time to fatten up on information? J Biol Chem 286,

11883-11889.

Brown, M.K., and Naidoo, N. (2012). The endoplasmic reticulum stress response

in aging and age-related diseases. Front Physiol 3, 263.

REFERENCES | 115

Browning, J.D., and Horton, J.D. (2004). Molecular mediators of hepatic steatosis

and liver injury. Journal of Clinical Investigation 114, 147-152.

Burns, D.M., and Richter, J.D. (2008). CPEB regulation of human cellular

senescence, energy metabolism, and p53 mRNA translation. Genes Dev 22, 3449-

3460.

Calderone, V., Gallego, J., Fernandez-Miranda, G., Garcia-Pras, E., Maillo, C.,

Berzigotti, A., Mejias, M., Bava, F.-A., Angulo-Urarte, A., Graupera, M., et al. (2016).

Sequential Functions of CPEB1 and CPEB4 Regulate Pathologic Expression of

Vascular Endothelial Growth Factor and Angiogenesis in Chronic Liver Disease.

Gastroenterology 150, 982-997.e930.

Calfon, M., Zeng, H., Urano, F., Till, J.H., Hubbard, S.R., Harding, H.P., Clark,

S.G., and Ron, D. (2002). IRE1 couples endoplasmic reticulum load to secretory

capacity by processing the XBP-1 mRNA. Nature 415, 92-96.

Chang, Y.W., and Huang, Y.S. (2014). Arsenite-activated JNK signaling enhances

CPEB4-Vinexin interaction to facilitate stress granule assembly and cell survival.

PLoS One 9, e107961.

Chen, P.J., and Huang, Y.S. (2012). CPEB2-eEF2 interaction impedes HIF-1alpha

RNA translation. EMBO J 31, 959-971.

Chen, W., Hu, Z., Li, X.Z., Li, J.L., Xu, X.K., Li, H.G., Liu, Y., Liu, B.H., Jia, W.H.,

and Li, F.C. (2015). CPEB4 interacts with Vimentin and involves in progressive

features and poor prognosis of patients with astrocytic tumors. Tumour Biol.

Chitnis, N., Pytel, D., and Diehl, J.A. (2013). UPR-inducible miRNAs contribute to

stressful situations. Trends in Biochemical Sciences 38, 447-452.

Cho, H., Zhao, X., Hatori, M., Yu, R.T., Barish, G.D., Lam, M.T., Chong, L.W.,

116 | REFERENCES

DiTacchio, L., Atkins, A.R., Glass, C.K., et al. (2012). Regulation of circadian

behaviour and metabolism by REV-ERB-alpha and REV-ERB-beta. Nature 485,

123-127.

Clarke, H.J., Chambers, J.E., Liniker, E., and Marciniak, S.J. (2014). Endoplasmic

reticulum stress in malignancy. Cancer Cell 25, 563-573.

Cnop, M., Foufelle, F., and Velloso, L.A. (2012). Endoplasmic reticulum stress,

obesity and diabetes. Trends in Molecular Medicine 18, 59-68.

Comuzzie, A.G., Cole, S.A., Laston, S.L., Voruganti, V.S., Haack, K., Gibbs, R.A.,

and Butte, N.F. (2012). Novel genetic loci identified for the pathophysiology of

childhood obesity in the Hispanic population. PLoS One 7, e51954.

Cretenet, G., Le Clech, M., and Gachon, F. (2010). Circadian clock-coordinated 12

Hr period rhythmic activation of the IRE1alpha pathway controls lipid metabolism

in mouse liver. Cell Metab 11, 47-57.

Crick, F. (1970). Central Dogma of Molecular Biology. Nature 227, 561-563.

Crick, F.H.C. (1958). On protein synthesis. Symp Soc Exp biol XII, 138.163.

Deng, J., Liu, S., Zou, L., Xu, C., Geng, B., and Xu, G. (2012). Lipolysis Response

to Endoplasmic Reticulum Stress in Adipose Cells. Journal of Biological Chemistry

287, 6240-6249.

Di Giammartino, D.C., Nishida, K., and Manley, J.L. (2011). Mechanisms and

consequences of alternative polyadenylation. Mol Cell 43, 853-866.

Douris, N., Kojima, S., Pan, X., Lerch-Gaggl, A.F., Duong, S.Q., Hussain, M.M.,

and Green, C.B. (2011). Nocturnin regulates circadian trafficking of dietary lipid in

intestinal enterocytes. Curr Biol 21, 1347-1355.

REFERENCES | 117

Drisaldi, B., Colnaghi, L., Fioriti, L., Rao, N., Myers, C., Snyder, A.M., Metzger,

D.J., Tarasoff, J., Konstantinov, E., Fraser, P.E., et al. (2015). SUMOylation Is an

Inhibitory Constraint that Regulates the Prion-like Aggregation and Activity of

CPEB3. Cell reports 11, 1694-1702.

Eckel-Mahan, K.L., Patel, V.R., Mohney, R.P., Vignola, K.S., Baldi, P., and Sassone-

Corsi, P. (2012). Coordination of the transcriptome and metabolome by the

circadian clock. Proc Natl Acad Sci U S A 109, 5541-5546.

Elkon, R., Ugalde, A.P., and Agami, R. (2013). Alternative cleavage and

polyadenylation: extent, regulation and function. Nat Rev Genet 14, 496-506.

Fabbrini, E., Sullivan, S., and Klein, S. (2010). Obesity and nonalcoholic fatty liver

disease: biochemical, metabolic, and clinical implications. Hepatology 51, 679-689.

Feldstein, A. (2010). Novel insights into the pathophysiology of nonalcoholic fatty

liver disease. Semin Liver Dis 30, 393-401.

Fernandez-Miranda, G., and Mendez, R. (2012). The CPEB-family of proteins,

translational control in senescence and cancer. Ageing research reviews 11, 460-

472.

Fioriti, L., Myers, C., Huang, Y.Y., Li, X., Stephan, J.S., Trifilieff, P., Colnaghi,

L., Kosmidis, S., Drisaldi, B., Pavlopoulos, E., et al. (2015). The Persistence of

Hippocampal-Based Memory Requires Protein Synthesis Mediated by the Prion-

like Protein CPEB3. Neuron 86, 1433-1448.

Fong, S.W., Lin, H.C., Wu, M.F., Chen, C.C., and Huang, Y.S. (2016). CPEB3

Deficiency Elevates TRPV1 Expression in Dorsal Root Ganglia Neurons to

Potentiate Thermosensation. PLoS One 11, e0148491.

Fu, S., Watkins, S.M., and Hotamisligil, G.S. (2012). The role of endoplasmic

118 | REFERENCES

reticulum in hepatic lipid homeostasis and stress signaling. Cell Metab 15, 623-634.

Fu, Y., Dominissini, D., Rechavi, G., and He, C. (2014). Gene expression regulation

mediated through reversible m(6)A RNA methylation. Nat Rev Genet 15, 293-306.

Gachon, F., and Bonnefont, X. (2010). Circadian clock-coordinated hepatic lipid

metabolism: only transcriptional regulation? Aging 2, 101-106.

Galabru, J., Katze, M.G., Robert, N., and Hovanessian, A.G. (1989). The binding

of double-stranded RNA and adenovirus VAI RNA to the interferon-induced

protein kinase. Eur J Biochem 178, 581-589.

Gao, X., Wan, J., Liu, B., Ma, M., Shen, B., and Qian, S.B. (2015). Quantitative

profiling of initiating ribosomes in vivo. Nat Methods 12, 147-153.

Garbarino-Pico, E., and Green, C.B. (2007). Posttranscriptional regulation of

mammalian circadian clock output. Cold Spring Harb Symp Quant Biol 72, 145-

156.

Garcia-Pras, E., Gallego, J., Coch, L., Mejias, M., Fernandez-Miranda, G., Pardal,

R., Bosch, J., Mendez, R., and Fernandez, M. (2016). Role and therapeutic potential

of vascular stem/progenitor cells in pathological neovascularisation during chronic

portal hypertension. Gut.

Gentile, C.L., Frye, M., and Pagliassotti, M.J. (2011). Endoplasmic reticulum stress

and the unfolded protein response in nonalcoholic fatty liver disease. Antioxid

Redox Signal 15, 505-521.

Gerstberger, S., Hafner, M., and Tuschl, T. (2014). A census of human RNA-

binding proteins. Nat Rev Genet 15, 829-845.

Giangarra, V., Igea, A., Castellazzi, C.L., Bava, F.A., and Mendez, R. (2015). Global

Analysis of CPEBs Reveals Sequential and Non-Redundant Functions in Mitotic

REFERENCES | 119

Cell Cycle. PLoS One 10, e0138794.

Glisovic, T., Bachorik, J.L., Yong, J., and Dreyfuss, G. (2008). RNA-binding proteins

and post-transcriptional gene regulation. FEBS Lett 582, 1977-1986.

Gray, N.K., and Wickens, M. (1998). Control of translation initiation in animals.

Annual review of cell and developmental biology 14, 399-458.

Hagele, S., Kuhn, U., Boning, M., and Katschinski, D.M. (2009). Cytoplasmic

polyadenylation-element-binding protein (CPEB)1 and 2 bind to the HIF-1alpha

mRNA 3’-UTR and modulate HIF-1alpha protein expression. Biochem J 417, 235-

246.

Hake, L.E., Mendez, R., and Richter, J.D. (1998). Specificity of RNA binding by

CPEB: requirement for RNA recognition motifs and a novel zinc finger. Mol Cell

Biol 18, 685-693.

Halbeisen, R.E., Galgano, A., Scherrer, T., and Gerber, A.P. (2008). Post-

transcriptional gene regulation: from genome-wide studies to principles. Cell Mol

Life Sci 65, 798-813.

Harding, H., Zhang, Y., Bertolotti, A., Zeng, H., and Ron, D. (2000). Perk is

essential for translational regulation and cell survival during the unfolded protein

response. Mol Cell 5, 897-904.

Harding, H.P., Zhang, Y., Zeng, H., Novoa, I., Lu, P.D., Calfon, M., Sadri, N., Yun,

C., Popko, B., Paules, R., et al. (2003). An Integrated Stress Response Regulates

Amino Acid Metabolism and Resistance to Oxidative Stress. Molecular Cell 11,

619-633.

Harding, H.P., Zhang, Y. & Ron, D. (1999). Protein translation and folding are

coupled by an endoplasmic-reticulum-resident kinase. Nature 397, 271-274.

120 | REFERENCES

Heid, I.M., Jackson, A.U., Randall, J.C., Winkler, T.W., Qi, L., Steinthorsdottir,

V., Thorleifsson, G., Zillikens, M.C., Speliotes, E.K., Magi, R., et al. (2010). Meta-

analysis identifies 13 new loci associated with waist-hip ratio and reveals sexual

dimorphism in the genetic basis of fat distribution. Nat Genet 42, 949-960.

Hollien, J., Lin, J.H., Li, H., Stevens, N., Walter, P., and Weissman, J.S. (2009).

Regulated Ire1-dependent decay of messenger RNAs in mammalian cells. J Cell

Biol 186, 323-331.

Hotamisligil, G.S. (2010). Endoplasmic Reticulum Stress and the Inflammatory

Basis of Metabolic Disease. Cell 140, 900-917.

Hu, W., Yang, Y., Xi, S., Sai, K., Su, D., Zhang, X., Lin, S., and Zeng, J. (2015).

Expression of CPEB4 in Human Glioma and Its Correlations With Prognosis.

Medicine (Baltimore) 94, 979.

Hu, W., Yuan, B., and Lodish, H.F. (2014). Cpeb4-mediated translational regulatory

circuitry controls terminal erythroid differentiation. Dev Cell 30, 660-672.

Huang, Y.S., Carson, J.H., Barbarese, E., and Richter, J.D. (2003). Facilitation of

dendritic mRNA transport by CPEB. Genes Dev 17, 638-653.

Huang, Y.S., Kan, M.C., Lin, C.L., and Richter, J.D. (2006). CPEB3 and CPEB4 in

neurons: analysis of RNA-binding specificity and translational control of AMPA

receptor GluR2 mRNA. EMBO J 25, 4865-4876.

Hummasti, S., and Hotamisligil, G.S. (2010). Endoplasmic reticulum stress and

inflammation in obesity and diabetes. Circ Res 107, 579-591.

Igea, A., and Mendez, R. (2010). Meiosis requires a translational positive loop

where CPEB1 ensues its replacement by CPEB4. EMBO J 29, 2182-2193.

Itoh, N., and Okamoto, H. (1980). Translational control of proinsulin synthesis by

REFERENCES | 121

glucose. Nature 283, 100-102.

Ivshina, M., Alexandrov, I.M., Vertii, A., Doxsey, S., and Richter, J.D. (2015). CPEB

regulation of TAK1 synthesis mediates cytokine production and the inflammatory

immune response. Mol Cell Biol 35, 610-618.

Ivshina, M., Lasko, P., and Richter, J.D. (2014). Cytoplasmic polyadenylation

element binding proteins in development, health, and disease. Annual review of

cell and developmental biology 30, 393-415.

Jackson, R.J., Hellen, C.U., and Pestova, T.V. (2010). The mechanism of eukaryotic

translation initiation and principles of its regulation. Nat Rev Mol Cell Biol 11,

113-127.

Jacob, F., and Monod, J. (1961). Genetic regulatory mechanisms in the synthesis of

proteins. J Mol Biol 3, 318-356.

Janich, P., Arpat, A.B., Castelo-Szekely, V., Lopes, M., and Gatfield, D. (2015).

Ribosome profiling reveals the rhythmic liver translatome and circadian clock

regulation by upstream open reading frames. Genome Res 25, 1848-1859.

Jostins, L., Ripke, S., Weersma, R.K., Duerr, R.H., McGovern, D.P., Hui, K.Y.,

Lee, J.C., Schumm, L.P., Sharma, Y., Anderson, C.A., et al. (2012). Host-microbe

interactions have shaped the genetic architecture of inflammatory bowel disease.

Nature 491, 119-124.

JS., O.N., and Reddy, A. (2011). Circadian Clocks in Human Red Blood Cells.

Nature 469, 498-503.

Kaczmarczyk, L., Labrie-Dion, E., Sehgal, K., Sylvester, M., Skubal, M., Josten,

M., Steinhauser, C., De Koninck, P., and Theis, M. (2016). New Phosphospecific

Antibody Reveals Isoform-Specific Phosphorylation of CPEB3 Protein. PLoS

122 | REFERENCES

One 11.

Kan, M.C., Oruganty-Das, A., Cooper-Morgan, A., Jin, G., Swanger, S.A., Bassell,

G.J., Florman, H., van Leyen, K., and Richter, J.D. (2010). CPEB4 is a cell survival

protein retained in the nucleus upon ischemia or endoplasmic reticulum calcium

depletion. Mol Cell Biol 30, 5658-5671.

Kaufman, R.J. (2002). Orchestrating the unfolded protein response in health and

disease. J Clin Invest 110, 1389-1398.

Keene, J.D. (2007). RNA regulons: coordination of post-transcriptional events.

Nat Rev Genet 8, 533-543.

Khan, Z., Ford, M.J., Cusanovich, D.A., Mitrano, A., Pritchard, J.K., and Gilad, Y.

(2013). Primate transcript and protein expression levels evolve under compensatory

selection pressures. Science 342, 1100-1104.

Kim, J.H., and Richter, J.D. (2006). Opposing polymerase-deadenylase activities

regulate cytoplasmic polyadenylation. Mol Cell 24, 173-183.

Kojima, S., Gendreau, K.L., Sher-Chen, E.L., Gao, P., and Green, C.B. (2015).

Changes in poly(A) tail length dynamics from the loss of the circadian deadenylase

Nocturnin. Sci Rep 5, 17059.

Kojima, S., Sher-Chen, E.L., and Green, C.B. (2012). Circadian control of mRNA

polyadenylation dynamics regulates rhythmic protein expression. Genes &

Development 26, 2724-2736.

Kornberg, D. (1999). Eukaryotic transcriptional control. Trends in Genetics 15,

46-49.

Lackner, D.H., and Bähler, J. (2008). Translational Control of Gene Expression:

from transcripts to transcriptomes. 271, 199-251.

REFERENCES | 123

Lee, A., Scapa, E., Cohen, D., and Glimcher, L. (2008). Regulation of hepatic

lipogenesis by the transcription factor XBP1. Science 320, 1492-1496.

Lee, A.H., Iwakoshi, N.N., and Glimcher, L.H. (2003). XBP-1 Regulates a Subset

of Endoplasmic Reticulum Resident Chaperone Genes in the Unfolded Protein

Response. Molecular and Cellular Biology 23, 7448-7459.

Lin, C.L., Huang, Y.T., and Richter, J.D. (2012). Transient CPEB dimerization and

translational control. RNA 18, 1050-1061.

Lu, L., Han, A.P., and Chen, J.J. (2001). Translation initiation control by heme-

regulated eukaryotic initiation factor 2alpha kinase in erythroid cells under

cytoplasmic stresses. Mol Cell Biol 21, 7971-7980.

Maerkeen, K. (2010). Tauroursodeoxycholic Acid May Improve Liver and Muscle

but Not Adipose Tissue Insulin Sensitivity in Obese Men and Women. Diabetes

59, 1899-1905.

Malhi, H., and Kaufman, R.J. (2011). Endoplasmic reticulum stress in liver disease.

J Hepatol 54, 795-809.

Malhotra, J.D., and Kaufman, R.J. (2011). ER stress and its functional link to

mitochondria: role in cell survival and death. Cold Spring Harb Perspect Biol 3.

Mauvoisin, D., Wang, J., Jouffe, C., Martin, E., Atger, F., Waridel, P., Quadroni,

M., Gachon, F., and Naef, F. (2014). Circadian clock-dependent and -independent

rhythmic proteomes implement distinct diurnal functions in mouse liver. Proc Natl

Acad Sci U S A 111, 167-172.

Mead, S., Uphill, J., Beck, J., Poulter, M., Campbell, T., Lowe, J., Adamson, G.,

Hummerich, H., Klopp, N., Ruckert, I.M., et al. (2012). Genome-wide association

study in multiple human prion diseases suggests genetic risk factors additional to

124 | REFERENCES

PRNP. Hum Mol Genet 21, 1897-1906.

Mendez, R., Hake, L.E., Andresson, T., Littlepage, L.E., Ruderman, J.V., and

Richter, J.D. (2000a). Phosphorylation of CPE binding factor by Eg2 regulates

translation of c-mos mRNA. Nature 404, 302-307.

Mendez, R., Murthy, K., Ryan, K., Manley, J., and Richter, J. (2000b). Phosphorylation

of CPEB by Eg2 mediates the recruitment of CPSF into an active cytoplasmic

polyadenylation complex. Mol Cell 6, 1253-1259.

Merkel, D.J., Wells, S.B., Hilburn, B.C., Elazzouzi, F., Perez-Alvarado, G.C., and

Lee, B.M. (2013). The C-terminal region of cytoplasmic polyadenylation element

binding protein is a ZZ domain with potential for protein-protein interactions. J

Mol Biol 425, 2015-2026.

Michelotti, G.A., Machado, M.V., and Diehl, A.M. (2013). NAFLD, NASH and

liver cancer. Nat Rev Gastroenterol Hepatol 10, 656-665.

Minshall, N., Reiter, M.H., Weil, D., and Standart, N. (2007). CPEB interacts with

an ovary-specific eIF4E and 4E-T in early Xenopus oocytes. J Biol Chem 282,

37389-37401.

Mohrin M, Shin J, Liu Y, Brown K, Luo H, Xi Y, Haynes CM, and D., C. (2015).

Stem cell aging. A mitochondrial UPR-mediated metabolic checkpoint regulates

hematopoietic stem cell aging. Science 347, 1374-1377.

Moore, M.J. (2005). From birth to death: the complex lives of eukaryotic mRNAs.

Science 309, 1514-1518.

Nagaoka, K., Fujii, K., Zhang, H., Usuda, K., Watanabe, G., Ivshina, M., and

Richter, J.D. (2015). CPEB1 mediates epithelial-to-mesenchyme transition and

breast cancer metastasis. Oncogene.

REFERENCES | 125

Nagaoka, K., Udagawa, T., and Richter, J.D. (2012). CPEB-mediated ZO-1 mRNA

localization is required for epithelial tight-junction assembly and cell polarity. Nat

Commun 3, 675.

Novoa, I., Gallego, J., Ferreira, P.G., and Mendez, R. (2010). Mitotic cell-cycle

progression is regulated by CPEB1 and CPEB4-dependent translational control.

Nature Cell Biology 12, 447-456.

Oakes, S.A., and Papa, F.R. (2015). The role of endoplasmic reticulum stress in

human pathology. Annu Rev Pathol 10, 173-194.

Orphanides, G., and Reinberg, D. (2002). A unified theory of gene expression. Cell

108, 439-451.

Ortiz-Zapater, E., Pineda, D., Martinez-Bosch, N., Fernandez-Miranda, G., Iglesias,

M., Alameda, F., Moreno, M., Eliscovich, C., Eyras, E., Real, F.X., et al. (2012). Key

contribution of CPEB4-mediated translational control to cancer progression. Nat

Med 18, 83-90.

Ota, T., Gayet, C., and Ginsberg, H.N. (2008). Inhibition of apolipoprotein B100

secretion by lipid-induced hepatic endoplasmic reticulum stress in rodents. J Clin

Invest 118, 316-332.

Ozcan, U., Cao, Q., Yilmaz, E., Lee, A., Iwakoshi, N., Ozdelen, E., Tuncman, G.,

Görgün, C., Glimcher, L., and Hotamisligil, G. (2004). Endoplasmic Reticulum

Stress Links Obesity, Insulin Action, and Type 2 Diabetes. Science 306, 457-461.

Ozcan, U., Yilmaz, E., Ozcan, L., Furuhashi, M., Vaillancourt, E., Smith, R.O.,

Görgün, C.Z., and Hotamisligil, G.S. (2006). Chemical Chaperones Reduce ER

Stress and Restore Glucose Homeostasis in a Mouse Model of Type 2 Diabetes.

Science 313, 1137-1140.

126 | REFERENCES

Papasaikas, P., and Valcarcel, J. (2016). The Spliceosome: The Ultimate RNA

Chaperone and Sculptor. Trends Biochem Sci 41, 33-45.

Pavitt, G.D., and Ron, D. (2012). New insights into translational regulation in the

endoplasmic reticulum unfolded protein response. Cold Spring Harb Perspect Biol

4.

Pavlopoulos, E., Trifilieff, P., Chevaleyre, V., Fioriti, L., Zairis, S., Pagano, A.,

Malleret, G., and Kandel, E.R. (2011). Neuralized1 Activates CPEB3: A Function

for Nonproteolytic Ubiquitin in Synaptic Plasticity and Memory Storage. Cell 147,

1369-1383.

Peek, C.B., Affinati, A.H., Ramsey, K.M., Kuo, H.Y., Yu, W., Sena, L.A., Ilkayeva,

O., Marcheva, B., Kobayashi, Y., Omura, C., et al. (2013). Circadian Clock NAD+

Cycle Drives Mitochondrial Oxidative Metabolism in Mice. Science 342.

Peng, W., Nan, Z., Liu, Y., Shen, H., Lin, C., Lin, L., and Yuan, B. (2014). Dendritic

cells transduced with CPEB4 induced antitumor immune response. Exp Mol

Pathol 97, 273-278.

Pique, M., Lopez, J.M., Foissac, S., Guigo, R., and Mendez, R. (2008). A combinatorial

code for CPE-mediated translational control. Cell 132, 434-448.

Postic, C., and Girard, J. (2008). Contribution of de novo fatty acid synthesis to

hepatic steatosis and insulin resistance: lessons from genetically engineered mice. J

Clin Invest 118, 829-838.

Postic, C., Shiota, M., Niswender, K.D., Jetton, T.L., Chen, Y., Moates, J.M., Shelton,

K.D., Lindner, J., Cherrington, A.D., and Magnuson, M.A. (1999). Dual roles for

glucokinase in glucose homeostasis as determined by liver and pancreatic beta cell-

specific gene knock-outs using Cre recombinase. J Biol Chem 274, 305-315.

REFERENCES | 127

Proud, C.G. (2002). Regulation of mammalian translation factors by nutrients. Eur

J Biochem 269, 5338-5349.

R., L., and AJ., S. (2013). The global NAFLD epidemic. Nat Rev Gastroenterol

Hepatol 10, 686-690.

Raabe, M., Veniant, M.M., Sullivan, M.A., Zlot, C.H., Bjorkegren, J., Nielsen,

L.B., Wong, J.S., Hamilton, R.L., and Young, S.G. (1999). Analysis of the role of

microsomal triglyceride transfer protein in the liver of tissue-specific knockout

mice. J Clin Invest 103, 1287-1298.

Rao, M.S., and Reddy, J.K. (2001). Peroxisomal β-Oxidation and Steatohepatitis.

Semin Liver Dis 21, 43-55.

Reid, D.W., Chen, Q., Tay, A.S., Shenolikar, S., and Nicchitta, C.V. (2014). The

unfolded protein response triggers selective mRNA release from the endoplasmic

reticulum. Cell 158, 1362-1374.

Reimold AM, Iwakoshi NN, Manis J, Vallabhajosyula P, Szomolanyi-Tsuda E,

Gravallese EM, Friend D, Grusby MJ, Alt F, and LH., G. (2001). Plasma cell

differentiation requires the transcription factor XBP-1. Nature 412, 300-307.

Reinke, H., and Asher, G. (2016). Circadian Clock Control of Liver Metabolic

Functions. Gastroenterology 150, 574-580.

Richter, J.D., and Coller, J. (2015). Pausing on Polyribosomes: Make Way for

Elongation in Translational Control. Cell 163, 292-300.

Richter, J.D., and Sonenberg, N. (2005). Regulation of cap-dependent translation

by eIF4E inhibitory proteins. Nature 433, 477-480.

Ron, D., and Walter, P. (2007). Signal integration in the endoplasmic reticulum

unfolded protein response. Nature Reviews Molecular Cell Biology 8, 519-529.

128 | REFERENCES

Rutkowski, D.T., Arnold, S.M., Miller, C.N., Wu, J., Li, J., Gunnison, K.M., Mori, K.,

Sadighi Akha, A.A., Raden, D., and Kaufman, R.J. (2006). Adaptation to ER stress

is mediated by differential stabilities of pro-survival and pro-apoptotic mRNAs

and proteins. PLoS Biol 4, e374.

Rutkowski, D.T., Wu, J., Back, S.H., Callaghan, M.U., Ferris, S.P., Iqbal, J., Clark, R.,

Miao, H., Hassler, J.R., Fornek, J., et al. (2008). UPR pathways combine to prevent

hepatic steatosis caused by ER stress-mediated suppression of transcriptional

master regulators. Dev Cell 15, 829-840.

Sachs, A.B., Sarnow, P., and Hentze, M.W. (1997). Starting at the beginning, middle,

and end: translation initiation in eukaryotes. Cell 89, 831-838.

Sagara, Y., and Inesi, G. (1991). Inhibition of the sarcoplasmic reticulum Ca2+

transport ATPase by thapsigargin at subnanomolar concentrations. J Biol Chem

266, 13503-13506.

Scheuner D, Song B, McEwen E, Liu C, Laybutt R, Gillespie P, Saunders T, Bonner-

Weir S, and RJ., K. (2001). Translational control is required for the unfolded protein

response and in vivo glucose homeostasis. Mol Cell 7, 1165-1176.

Schwanhausser, B., Busse, D., Li, N., Dittmar, G., Schuchhardt, J., Wolf, J., Chen,

W., and Selbach, M. (2011). Global quantification of mammalian gene expression

control. Nature 473, 337-342.

Shen, J., Chen, X., Hendershot, L., and Prywes, R. (2002). ER stress regulation of

ATF6 localization by dissociation of BiP/GRP78 binding and unmasking of Golgi

localization signals. Dev Cell 3, 99-111.

Shi, Y., Vattem, K.M., Sood, R., An, J., Liang, J., Stramm, L., and Wek, R.C. (1998).

Identification and characterization of pancreatic eukaryotic initiation factor 2

alpha-subunit kinase, PEK, involved in translational control. Mol Cell Biol 18,

REFERENCES | 129

7499-7509.

Singh, G., Pratt, G., Yeo, G.W., and Moore, M.J. (2015). The Clothes Make the

mRNA: Past and Present Trends in mRNP Fashion. Annu Rev Biochem 84, 325-

354.

So, J. (2012). Silencing of lipid metabolism genes through IRE1α-mediated mRNA

decay lowers plasma lipids in mice. Cell Metab 16, 487-499.

Sonenberg, N., and Hinnebusch, A.G. (2009). Regulation of translation initiation

in eukaryotes: mechanisms and biological targets. Cell 136, 731-745.

Stebbins-Boaz, B., Cao, Q., de Moor, C.H., Mendez, R., and Richter, J.D. (1999).

Maskin is a CPEB-associated factor that transiently interacts with elF-4E. Mol Cell

4, 1017-1027.

Stephan, J.S., Fioriti, L., Lamba, N., Colnaghi, L., Karl, K., Derkatch, I.L., and

Kandel, E.R. (2015). The CPEB3 Protein Is a Functional Prion that Interacts with

the Actin Cytoskeleton. Cell reports 11, 1772-1785.

Sun, H.T., Wen, X., Han, T., Liu, Z.H., Li, S.B., Wang, J.G., and Liu, X.P. (2015).

Expression of CPEB4 in invasive ductal breast carcinoma and its prognostic

significance. Onco Targets Ther 8, 3499-3506.

Tabas, I., and Ron, D. (2011). Integrating the mechanisms of apoptosis induced by

endoplasmic reticulum stress. Nat Cell Biol 13, 184-190.

Tarun, S.Z., Jr., and Sachs, A.B. (1995). A common function for mRNA 5’ and 3’

ends in translation initiation in yeast. Genes Dev 9, 2997-3007.

Theis, M., Si, K., and Kandel, E.R. (2003). Two previously undescribed members

of the mouse CPEB family of genes and their inducible expression in the principal

cell layers of the hippocampus. Proc Natl Acad Sci U S A 100.

130 | REFERENCES

Topisirovic, I. (2011). Cap and cap-binding proteins in the control of gene

expression. Wiley Interdiscip Rev RNA 2, 277-298.

Tsai, L.Y., Chang, Y.W., Lin, P.Y., Chou, H.J., Liu, T.J., Lee, P.T., Huang, W.H., Tsou,

Y.L., and Huang, Y.S. (2013). CPEB4 knockout mice exhibit normal hippocampus-

related synaptic plasticity and memory. PLoS One 8.

Udagawa, T., Swanger, S.A., Takeuchi, K., Kim, J.H., Nalavadi, V., Shin, J., Lorenz,

L.J., Zukin, R.S., Bassell, G.J., and Richter, J.D. (2012). Bidirectional control of

mRNA translation and synaptic plasticity by the cytoplasmic polyadenylation

complex. Mol Cell 47, 253-266.

van Galen, P., Kreso, A., Mbong, N., Kent, D.G., Fitzmaurice, T., Chambers, J.E.,

Xie, S., Laurenti, E., Hermans, K., Eppert, K., et al. (2014). The unfolded protein

response governs integrity of the haematopoietic stem-cell pool during stress.

Nature 510, 268-272.

Vattem, K.M., and Wek, R.C. (2004). Reinitiation involving upstream ORFs

regulates ATF4 mRNA translation in mammalian cells. Proceedings of the National

Academy of Sciences 101, 11269-11274.

Vollmers, C., Gill, S., DiTacchio, L., Pulivarthy, S.R., Le, H.D., and Panda, S. (2009).

Time of feeding and the intrinsic circadian clock drive rhythms in hepatic gene

expression. Proc Natl Acad Sci U S A 106.

Volmer, R., and Ron, D. (2015). Lipid-dependent regulation of the unfolded protein

response. Curr Opin Cell Biol 33, 67-73.

Walter, P., and Ron, D. (2012). The Unfolded Protein Response: From Stress

Pathway to Homeostatic Regulation.

Wang, L., Zeng, X., Ryoo, H.D., and Jasper, H. (2014). Integration of UPRER and

REFERENCES | 131

oxidative stress signaling in the control of intestinal stem cell proliferation. PLoS

Genet 10.

Wang, M., and Kaufman, R.J. (2016). Protein misfolding in the endoplasmic

reticulum as a conduit to human disease. Nature 529, 326-335.

Wang, S. (2012). IRE1α-XBP1s induces PDI expression to increase MTP activity

for hepatic VLDL assembly and lipid homeostasis. Cell Metab 16, 473-486.

Wang, S., and Kaufman, R.J. (2014). How does protein misfolding in the endoplasmic

reticulum affect lipid metabolism in the liver? Curr Opin Lipidol 25, 125-132.

Wang, X.P., and Cooper, N.G. (2010). Comparative in silico analyses of cpeb1-4

with functional predictions. Bioinform Biol Insights 4, 61-83.

Weill, L., Belloc, E., Bava, F.A., and Mendez, R. (2012). Translational control by

changes in poly(A) tail length: recycling mRNAs. Nat Struct Mol Biol 19, 577-585.

Weingarten-Gabbay, S., Elias-Kirma, S., Nir, R., Gritsenko, A.A., Stern-Ginossar,

N., Yakhini, Z., Weinberger, A., and Segal, E. (2016). Systematic discovery of cap-

independent translation sequences in human and viral genomes. Science 351.

Wek, R.C., Jiang, H.Y., and Anthony, T.G. (2006). Coping with stress- eIF2 kinases

and translational control. Biochem Soc Trans 34, 7-11.

Wijnen, H., and Young, M.W. (2006). Interplay of circadian clocks and metabolic

rhythms. Annual review of genetics 40, 409-448.

Wu, L., Candille, S.I., Choi, Y., Xie, D., Jiang, L., Li-Pook-Than, J., Tang, H., and

Snyder, M. (2013). Variation and genetic control of protein abundance in humans.

Nature 499, 79-82.

Yoshida, H., Matsui, T., Yamamoto, A., Okada, T., and Mori, K. (2001). XBP1

132 | REFERENCES

mRNA Is Induced by ATF6 and Spliced by IRE1 in Response to ER Stress to

Produce a Highly Active Transcription Factor. Cell 107, 881-891.

Zhang K1, W.H., Song B, Miller CN, Scheuner D, Kaufman RJ. (2005). The

unfolded protein response sensor IRE1alpha is required at 2 distinct steps in B cell

lymphopoiesis. J Clin Invest 115, 268-281

Zhong, X., Xiao, Y., Chen, C., Wei, X., Hu, C., Ling, X., and Liu, X. (2015).

MicroRNA-203-mediated posttranscriptional deregulation of CPEB4 contributes

to colorectal cancer progression. Biochem Biophys Res Commun 466, 206-213.

Appendix

134 | APPENDIX

symbol log2FC symbol log2FC symbol log2FCNpm3 3,69 Tat 2,65 Ptplb 2,47Brd3 3,53 Mex3c 2,64 Atp11b 2,47Pck1 3,47 Cmip 2,64 Dpyd 2,46

Cadm1 3,40 Snx9 2,64 Snrnp40 2,46Elovl2 3,37 Csgalnact2 2,63 Grb7 2,46Cpeb4 3,29 Ubl4 2,63 Ppil3 2,44Socs1 3,29 Slc10a3 2,63 Lif 2,44

Prpf40a 3,25 Wtap 2,62 Tank 2,43Cog6 3,15 Nfkbia 2,62 Akirin1 2,43Cstf1 3,08 Tsen15 2,61 Adk 2,43

I830012O16Rik 3,05 Tiparp 2,61 Ugdh 2,43Sqrdl 3,04 Tjp1 2,60 Phip 2,42

Gm15441 3,01 Dek 2,60 Smarca5 2,42Txnip 3,01 Slc35b3 2,59 Dr1 2,42

Tcp11l2 3,00 Cks1bGm6340 2,59 Slc2a2 2,41Hnrnpd 2,99 Gyk 2,58 Dmxl1 2,41

Xpot 2,97 Cmpk2 2,58 Pcgf5 2,41Cpsf1 2,95 Hdac1 2,58 l7Rn6 2,41Trp53 2,93 Wbp5 2,58 Krt5 2,40Clcc1 2,92 Mier3 2,58 Jak2 2,40

Gpsm2 2,92 Gbp3 2,57 Yaf2 2,40Sde2 2,90 Efna1 2,56 Rala 2,40Tap1 2,90 Sap130 2,56 Nedd1 2,39

Snrpb2 2,89 Atg5 2,56 Lsm3 2,39Tshz1 2,89 NA 2,55 Sf3a3 2,39Bhmt 2,87 Tgds 2,55 Vezf1 2,38Bcl3 2,87 Srek1ip1 2,55 Hmgb1 2,38

Ginm1 2,87 Nfkb1 2,55 Jag1 2,38Foxq1 2,86 Dnajc2 2,55 Clic1 2,37Mzt1 2,85 Psma1 2,55 Gjb2 2,37

Gabpa 2,84 Pprc1 2,55 Mitd1 2,37Josd2 2,84 Cdk4 2,54 Tbp 2,36

Psmd11 2,83 Psmb9 2,54 M6pr 2,36Med31 2,80 Enpp4 2,53 Cyp1a2 2,36Setd4 2,80 Efnb1 2,53 Rsad2 2,36Ier3ip1 2,80 Spryd7 2,53 Tgfbr2 2,35Hdhd2 2,80 00Rik 2,52 Rcn2 2,35

Gm4788 2,79 Scpep1 2,52 Slmap 2,34

Table 1. Top enriched candidates of the RIP-seq analysis.

APPENDIX | 135

Gtpbp10 2,79 Stx4a 2,52 Pex11a 2,34Nae1 2,79 Kdm6b 2,52 Uroc1 2,34Bcl10 2,78 Heatr5a 2,52 Serinc5 2,34

Tma7-psTma7 2,78 Slco1b2 2,52 Bud13 2,33Fam49b 2,77 Cnep1r1 2,51 Art3 2,32Sema3e 2,77 Sbds 2,51 Cxcl10 2,32Cldn25 2,74 Pcmtd2 2,51 Nipsnap3b 2,32Tpra1 2,73 Pdcl3 2,51 Hpd 2,32Acat1 2,72 Sacm1l 2,51 Sema4b 2,32

Kdm2a 2,72 Samd1 2,50 Mkl1 2,31Coq10b 2,72 Eny2 2,50 Kdm5a 2,31Lmnb1 2,71 Polr2d 2,49 Tomm22 2,31Arg1 2,71 H2afv 2,49 Rap2c 2,31

Cyp2c50 2,70 Herpud2 2,49 Apaf1 2,31Usb1 2,70 Ppif 2,49 Slc30a1 2,3100Rik 2,70 Git2 2,48 Ccar1 2,31Yipf4 2,68 Snx3 2,48 Cnot3 2,31

Abhd16a 2,68 Eif4a3Gm5576 2,48 Leng1 2,31Copb1 2,68 Ssr2 2,48 Slc25a32 2,30

F2r 2,68 Polr2hGm7511 2,48 Ifit3 2,30Tmem64 2,67 Vprbp 2,48 Il1rap 2,30

Naf1 2,67 Manf 2,48 Ctdp1 2,30Il1rn 2,66 Nampt 2,47 Arl6ip6 2,30Tcf12 2,65 00Rik 2,47 Wipi1 2,30

Tmem39a 2,30 Cpox 2,22 Sp1 2,14Fgfr2 2,29 Ocln 2,22 Zfyve16 2,14Oat 2,29 Snx7 2,21 Zfp422 2,14Eed 2,29 Mex3d 2,21 Rnf149 2,14

Zfp507 2,29 Kdelr2 2,20 Anp32a 2,14Cks2 2,29 Psma7 2,20 Snx1 2,14

Lrrc40 2,29 Rap1a 2,20 Top2b 2,14Ing3 2,29 D19Bwg1357e 2,20 Hspa13 2,14

Tma16 2,29 Mettl2 2,20 Metap2 2,13Atp2b1 2,29 Rcc1 2,20 Tor3a 2,13Sult1d1 2,29 Twsg1 2,20 Prkaa1 2,13

Rpia 2,28 Elf1 2,20 Katnbl1 2,13Rbm22 2,28 Naga 2,20 Dync1i2 2,13Ube2d1 2,28 Plagl1 2,19 Bbx 2,1300Rik 2,28 Lgals8 2,19 Cdkn1b 2,13Ppib 2,28 Kif21a 2,19 Smndc1 2,13

136 | APPENDIX

Gene symbol log2 FoldChange padjIrf7 -0,427 0,096Bst2 -0,430 0,044Mid1ip1 -0,449 0,096Isg15 -0,480 0,096Paqr7 -0,501 0,065Cyp2d40 -0,509 0,047Gbp6 -0,547 0,044Gck -0,889 1,13E-7

Table 2. mRNAs with reduced localization at the ER upon CPEB4 depletion